Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Biomedically relevant coordination chemistry of rhenium and the lanthanides Xu, Liang 2000

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

Item Metadata

Download

Media
831-ubc_2001-612066.pdf [ 4.39MB ]
Metadata
JSON: 831-1.0061239.json
JSON-LD: 831-1.0061239-ld.json
RDF/XML (Pretty): 831-1.0061239-rdf.xml
RDF/JSON: 831-1.0061239-rdf.json
Turtle: 831-1.0061239-turtle.txt
N-Triples: 831-1.0061239-rdf-ntriples.txt
Original Record: 831-1.0061239-source.json
Full Text
831-1.0061239-fulltext.txt
Citation
831-1.0061239.ris

Full Text

BIOMEDICALLY R E L E V A N T COORDINATION CHEMISTRY OF RHENIUM AND THE LANTHANIDES by Liang Xu B.Sc. The University of Victoria, 1995 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENT FOR THE DEGREE OF DOCTOR OF PHILOSOPHY In THE F A C U L T Y OF G R A D U A T E STUDIES (Department of chemistry) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA October 2000 © Liang Xu, 2000 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. The University of British Columbia Vancouver, Canada Department Date Pe-C . [2 . 2~OOfi DE-6 (2/88) ABSTRACT [ReO(ppme)X] (where ppme2" is 2,5-diazo-N,N'-dimethylhexyl-l,6-bis(phenylphosphinate), X = Bro.3Clo.7) has been synthesized via a substitution reaction and structurally characterized. [ReO(ppme)Cl] has been prepared by a reduction/complexation reaction from [NH4][Re04]. [ReO(ppme)Cl] reacts with thiocyanate and benzene thiolate forming [ReO(ppme)X] (X = "NCS, "SC6H5), but the one-pot synthesis of the respective ternary thiolate complexes from perrhenate was not successful. The reduction/complexation reaction of /?-methoxybenzenethiol, F^ppmeCL;, and perrhenate resulted in the formation of [H3ppme][ReO(SR)4], the reaction of which with ReO(ppme)Cl does not lead to [ReO(ppme)SR] in high yields. The oxorhenium(V) complexes with Hapmen (N,N'-bis(2-pyridylmethyl)ethylenediamine, an N4 ligand) and fl^bped (N,N'-bis(2-pyridylmethyl)ethylenediamine-N,N'-diacetic acid, a n ^ C h ligand) have been synthesized and structurally characterized. Along with [ReO(bbpen)]+ (where F^bbpen is N,N'-bis(2-hydroxybenzyl)-N,N'-bis(2-methylpyridyl)ethylenediamine, anN402 ligand) and [ReO(qtp)(OCH3)2]+ (where qtp, 2,2'-6',2"-6",2,"-quaterpyridine, has a n N 4 donor set), [ReO(H2pmen)Cl2]+ and [ReO(bped)]+ complexes share a rare seven-coordinate structure with a distorted pentagonal bipyramidal geometry, which represents a potentially general structural motif in Re v=0 complexes. The complexes retained their structure in solution. Chemistry of Re complexes with Fbpmen ligand has been further investigated. Complex cations [ReO(H2pmen)Cl2]+, [ReO(Hpmen)Cl]+ and [Re02(H2pmen)]+ are related by hydrolysis/HCl substitution. [ReO(Hpmen)Cl]+ was structurally characterized and found ii to contain a water-stable amido-Re bond. Dehydrogenation of the N-donor ligand from amine to imine with concomitant two-electron reduction of the Re center occurs readily in these systems. With suitable co-ligands such as 3-hydroxy-4-pyrones, ternary complexes [ReCl(L)(Ci4Hi4N4)][Re04] (where C i 4 H i 4 N 4 = H2pmen-4H and L = 3-hydroxy-4-pyronate anions) can be made from [NH4][Re04], H2pmen-4HC1 and pyrones in one-pot syntheses. These complexes may prove useful in nuclear medicine. H 5 X T (bis{[bis(carboxymethyl)amino]methyl}phosphinic acid) is an EDTA4"-like ligand containing an extra phosphinate group. [Co"(XT)]3", [Com(XT)]2" and a series of [Ln(XT)] ", complexes have been prepared. The phosphinate group is not coordinated in the Co complexes but is bound in the lanthanide complexes. Solid state and solution behaviours of Ln-XT species are consistent: both mono-protonated and non-protonated species have been found. Protonation of the metal complex does not lead to dissociation of a carboxylate, rather, the proton distributes around the molecular ion. The pM values of Ln-XT are comparable to those of Ln-EDTA but are higher than those of Ln-TMDTA. The inclusion of a phosphinate eases the selectivity of an EDTA-type ligand to late lanthanides. iii TABLE OF CONTENTS Abstract ii Table of Contents iv List of Tables vi List of Figures vii List of Abbreviations ix Acknowledgments xii Chapter 1 General Introduction 1 References 13 Chapter 2 [ReO(N 20 2)X] Complexes: 4+1? 18 2.1 Introduction 18 2.2 Experimental 20 2.3 Results and Discussion 25 2.4 Conclusions 30 2.5 References 31 Chapter 3 Seven-Coordinate [Re v ON 4 X 2 ] + Complexes (X = O and CI) 33 3.1 Introduction 33 3.2 Experimental 35 3.3 Results and Discussion 37 3.4 Conclusions 47 3.5 References 47 Chapter 4 Chemistry of Re with N,N'-Bis(2-pyridylmethyl)ethylenediamine 49 iv (H2pmen): Hydrolysis, Dehydrogenation, and Ternary Complexes 4.1 Introduction 49 4.2 Experimental 51 4.3 Results and Discussion 56 4.4 Conclusions 69 4.5 References 70 Chapter 5 Lanthanide Chemistry with 73 Bis {[bis(carboxymethyl)amino]methyl} phosphinate: What Does an Extra Phosphinate Group Do to EDTA? 5.1 Introduction 73 5.2 Experimental 75 5.3 Results and Discussion 81 5.4 Conclusions 92 5.5 References 92 Chapter 6 Conclusions and Further Thoughts 95 6.1 Rhenium Chemistry 95 6.1.1 Conclusions 95 6.1.2 Further thoughts 97 6.2 Lanthanide Chemistry 101 6.2.1 Conclusions 101 6.2.2 Further thoughts 101 6.3 References 104 Appendix 106 v List of Tables Table Title 2.1. Selected Bond Lengths (A) and Bond Angles (deg) 27 in [ReO(ppme)Br0.3Clo.7]. 3.1. Selected Bond Lengths (A) and Angles (deg) in the Three 42 Complex Cations. 3.2. 'H NMR Data for the Re(V) Complexes (Scheme 3.2). 46 4.1. Selected Bond Lengths (A) and Bond Angles (deg) in the 58 Complex Cations in 4.2 and 4.5. 5.1. Selected Bond Lengths (A) and Angles (deg) in [Co n(XT)] 3\ 5.1 5.2. Step-Wise Protonation Constants (25°C). 86 5.3. Logarithms of the Formation Constants for Complexation of 88 L n m with X T (T = 25°C and p. = 0.16 M NaCl). 5.4. pM Values of Relevant Systems ([L] = 10"5 M , [M] = 10"6 M, 90 pH = 7.4, T = 25°C). A l . Selected Crystallographic Data for [ReO(ppme)X] X = Br0.29/Cl0.7i- 106 A2. Selected Crystallographic Data for Seven-Coordinate Re v O Complexes. 107 A3. Selected Crystallographic Data for 4.2 and 4.5. 108 A4. Selected Crystallographic Data for K 3[Co(XT)]-2H 20. 109 vi List of Figures Figure Title 2.1. ORTEP Drawing of [ReO(ppme)Bro.3Clo.7]. Thermal Ellipsoids 26 for the Non-Hydrogen Atoms are Drawn at the 33% Probability Level. 3.1. Thermal Ellipsoid Drawing of the [ReO(H2pmen)Cl2]+ Cation in 38 [ReO(H2pmen)Cl2]Cl-2CH3OH Showing the Crystallographic Numbering. Thermal Ellipsoids for the Non-Hydrogen Atoms are Drawn at the 50% Probability Level. 3.2 Thermal Ellipsoid Drawing of the [ReO(bped)]+ Cation in 39 [ReO(bped)][Re04]-CH3OH Showing the Crystallographic Numbering. Thermal Ellipsoids for the Non-Hydrogen Atoms are Drawn at the 50% Probability Level. 3.3. Thermal Ellipsoid Drawing of the [ReO(bbpen)]+ Cation in 40 [ReO(bbpen)]PF6 Showing the Crystallographic Numbering. Thermal Ellipsoids for the Non-Hydrogen Atoms are Drawn at the 50% Probability Level. 3.4. Drawing of the Coordinating Atoms in the [ReO(H2pmen)Cl2]+, 41 [ReO(bped)]+ and [ReO(bbpen)]+ Cations. 4.1. ORTEP Drawing of the [ReOCl(Hpmen)]+Cation in 4.2, 57 Showing the Crystallographic Numbering. Thermal Ellipsoids For the Non-Hydrogen Atoms are Drawn at the 33% Probability Level. 4.2. J H N M R Spectrum of [ReOCl(Hpmen)][Re04], 4.2, in DMSO-d 6 . 59 4.3. 1HNMRSpectrumof[Re(Ci4Hi 6N 4)(NCS) 3], 4.4, in Acetone-^. 63 vii 4.4. ! H NMR Spectrum of [ReCl(ma)(Ci4H1 4N4)][Re04], 4.5, 64 in CDCI3/CD3OD. 4.5. ORTEP Drawing of the [ReCl(ma)(CHHi4N4)]+ Cation in 4.5, 66 Showing the Crystallographic Numbering. Thermal Ellipsoids for the Non-Hydrogen Atoms are Drawn at the 33% Probability Level. 5.1. Representative IR spectrum (K2[Gd(XT)](H20)4) of Ln-XT 77 and Ln-HXT Complexes. 5.2. ORTEP Drawing of the [Co"(XT)]3- Anion in K 3[Co(XT)]-2H 20. 82 The molecule lies on a crystallographic C 2 axis through Co(l) and P(l). Thermal Ellipsoids for the Non-Hydrogen Atoms are Drawn at the 33% Probability Level. 5.3. Speciation Diagram of the Gd-XT system ([Gd] - [XT] = 1 mM, 87 T = 25°C, n = 0.16 M NaCl). 5.4. Trends in pM values (See Table 5.4 for Conditions). 91 viii List of Abbreviations Abbreviation Meaning A angstrom, 1 x 10"10 metre Anal analytical P beta particles, overall stability constant of n stepwise equilibria (pn) B A M biologically active molecule B A T O boronic acid adducts of technetium dioximes BBB blood-brain-barrier B F C A bifunctional chelator Calcd calculated CDO cyclohexanedione dioxime cm"1 wavenumber(s) (reciprocal centimeter) 8 chemical shift in parts per million downfield from tetramethylsilane (lH NMR) or H3PO4 ( 3 1P NMR) d doublet (NMR) DMSO dimethylsulfoxide E A elemental analysis E C D ethylcysteine dimmer eV electron volts FAb fraction antigen binding g gram h hour(s) ix H2bbpen N,N'-bis(2-hydroxybenzyl)-N,N'-bis(2-methylpyridyl)ethylenediamine H2bped N,N' -bis(2-pyridylmethyl)ethylenediamine-N,N' -diacetic acid H 5 D T P A diethylenetriamine-N,N,N' ,N",N"-pentaacetic acid H4EDTA ethylenediamine-N,N,N' ,N' -tetraacetic acid HgEDTMP ethylenediamine-N,N,N' ,N' -tetrakis(methylenephosphonic acid) Hema ethylmaltol, 3-hydroxy-2-ethyl-4-pyrone Hfla flavonol, 3-hydroxyflavone H4HEDP 1,1 -hydroxyethylidenediphosphonic acid Hkoj kqjic acid, 5-hydroxy-2-hydroxymethyl-4-pyrone Hma maltol, 3-hydroxy-2-methyl-4-pyrone HMPAO 3,6,6,9-tetramethyl-4,8-diazaundecane-2,10-dione dioxime HPLC high performance liquid chromatography H2pmen N,N'-bis(2-pyridylmethyl)ethylenediamine H2ppme 2,5-diazo-N,N'-dimethylhexyl-l,6-bis(phenylphosphinic acid) H5TMDTA trimethylenedinitrilotetraacetic acid H 5 X T bis{[bis(carboxymethyl)amino]methyl}phosphinic acid HYNIC hydrazinonicotinic acid H Y P Y hydrazinopyridine Hz hertz IR infrared K stepwise stability constant L litre Ln lanthanides LSIMS liquid secondary ion mass spectrometry jo, micro (10"6) m milli-M molar (moles/litre for concentration); metal m/z mass-to-charge ratio (in mass spectrometry) MAb monoclonal antibody min minute(s) MRI magnetic resonance imaging v stretching frequency (IR) N M R nuclear magnetic resonance ORTEP Oak Ridge Thermal Ellipsoid Program PC personal computer pH negative log of the proton concentration or activity ppm parts per million py pyridyl qtp 2,2'-6\2"-6",2"'-quaterpyridine s second(s); singlet(NMR) T temperature t triplet ti/2 half-life tacn 1,4,7-triazacyclononane td triplet of doublets (NMR) TMS tetramethylsilane xi Acknowledgments I must acknowledge Professor Chris Orvig for his support, patience, and enthusiasm, as well as for allowing scientific freedom, for encouraging intellectual growth, and for indulging occasional (and at times wacky) creativity. I would like to thank Dr. Kathie Thompson for helpful discussions and encouragement. I would like to thank Dr. Brian Patrick, Dr. Victor Young, Dr. Maren Pink and the late Dr. Steve Rettig for X-ray structural determinations. Thanks to the Brits, the Europeans, the Boys, the Girls, and a few singularities who do not fit into any category, collectively known as the Orvig group, who made my time in Vancouver a memorable one. Special thanks go to Ika, Marco, Peter, Ashley, Song Bin, Barry, and Mike. Thanks to the departmental support staff. xii Chapter 1. General Introduction Much rhenium research has been fueled by interest in technetium, whose radionuclide ( 9 9 m Tc) is the workhorse of diagnostic nuclear medicine.1 ~9 Recently, however, the radioisotopes 1 8 6 Re (ti/2=90 h, E p m a x = 1.07 MeV) and 1 8 8 Re (ti/2=17 h, E p m a x = 2.11 MeV) have been investigated for potential use in therapeutic nuclear medicine: the use of their complexes to deliver ionizing radiation to cancerous tissues.4-6'8 For example, a 1 8 6 Re complex with hydroxyethylidene diphosphonate (HEDP) has been proven to be beneficial in clinical trials for the palliation of bone pain. 4' 6' 1 0 The effective range for 1 8 6 Re is 5 mm, suitable for small tumors, and that for 1 8 8 Re is 11 mm, useful for larger ones. 1 8 6 Re is i oo 188 2 generated by neutron irradiation of Re, and Re is readily available from a [ W O 4 ] " generator as [188Re04J". The availability of 1 8 8 Re contributes significantly to the recent attention paid to Re chemistry and its applications in nuclear medicine, and will attract more research in this area. Burgeoning from it, this thesis focuses mainly on the design and synthesis of rhenium complexes with biomedical relevance from the perspective of coordination chemistry. General necessities for all metal complexes as radiopharmaceuticals are the same: thermodynamic stability/kinetic inertness, target specificity, fast localization and clearance, and low toxicity; and for Tc and Re drugs, fast and easy synthesis from [MO4]". Similarities in Re and Tc chemistry mean that Re pharmaceuticals can be correlated to the well-established Tc counterparts, but the extent of this correlation is limited. While some complexes of Tc and Re may be similar, their redox and kinetic behaviours tend to be different, with Re complexes, generally speaking, being more inert and easily oxidized. The 1 implication of this difference is two-fold. First is a practical challenge of reducing perrhenate and complexing it to a suitable ligand in a reasonable time frame. Secondly, the biochemistry of Re and Tc complexes may be totally different. For example, many Re complexes whose Tc counterparts were shown to be easily demetallated by challenging ligands would probably survive biological metal scavengers long enough to be useful. The difficulty in using Re drugs arises from the demanding nature of therapeutic radiopharmaceuticals. An ideal therapeutic agent would quickly locate and concentrate on the cancerous foci with high specificity to deliver its radiation dose while producing minimal radiation to healthy tissues. While targetbackground ratio of a diagnostic Tc drug need only be a small integer for imaging purposes, for Re drugs, such a targetbackground ratio would mean that the majority of the (damaging) radiation dose is to healthy tissue. Thus the requirement for the drug to have high uptake and retention in the target (tumor) tissue, and fast excretion, is more stringent. This might be the reason that there are fewer commercial drugs in therapeutic than in diagnostic nuclear medicine, and the successful ones are simple complexes designed with a clear metabolic destination and fast clearance in mind. Thus anionic and hydrophilic complexes prevail. For example,13'I- sodium iodide is used for treatment of thyroid cancer, 8 9Sr- strontium chloride, 32P-sodium phosphate and Other phosphate complexes are used for the palliation of bone pain.4-6'8 Research on more sophisticated drugs is underway on both fronts, drawing correlations from established Tc systems, and on a more fundamental level, designing and developing new coordination complexes.4'6'8 In formulating a design strategy for Re complexes as radiopharmaceuticals, much can be learned from the evolution of the well-established Tc complexes in nuclear medicine. In 2 the early stages of Tc drug development, direct labeling prevailed. Pertechnetate was mixed with a ligand (e.g. HEDP) or other binding agents with or without a reductant.3'4 The simplest example is Tc-sulfur colloid, which is produced by sodium dithionite reduction of [TCO4]" in an acidic solution, and is used for liver imaging.3'4 The chemical nature of some of these materials has not been well-defined, even to this day. Active research in this area now focuses only on labeling macromolecules, i.e. monoclonial antibodies (MAb) or their fragments (FAb) . 1 1 - 1 3 Introduction of a coordinating metal often disrupts the native structures of the MAb/FAb and renders them less effective binders to their target molecules because a number of functional groups on the antibody, often serving structural roles, are coordinated to the metal. (It is generally thought that the metal disrupts disulfide linkages and binds to the resulting thiolates.)11 Therefore the major challenge facing a coordination chemist is how to effectively tag the metal onto MAb with minimum disruption to the structure of the latter. The use of chemically well-defined Tc complexes started with the "small molecule approach", in which a coordination complex was synthesized and then its in vivo biodistribution studied.3 Tc complexes can be synthesized both on a synthetic scale ( 9 9Tc) and on a tracer scale ( 9 9 mTc), and their chemistry can be correlated by, for example, HPLC. This correlation is not always possible considering that there is at least a 105 dilution factor between the two reactions conditions. After a coordination complex is defined, the ligand is often modified to alter the chemical and physical properties of the metal complex thereby tinkering with its biochemistry. Much Tc chemistry has been learned and many successful examples of this approach are on the market now.3 3 A very distinctive feature emerges from a number of successful Tc radiopharmaceuticals: it has been found that many of them are distributed by perfusion and retained by biomodification.3 For example, 9 9 m T c O - L , L - E C D (Neurolite™) and 9 9 m T c O -D , L - H M P A 0 1 4 (Ceretec™) are brain imaging agents (see Scheme 1.1 on next page). They are lipophilic, a property that enables them to cross the blood-brain-barrier; once they enter the brain they are stereospecifically hydrolyzed and their metabolites are retained in the brain. 1 5' 1 6 Cardiolite™, a lipophilic monocationic complex, is a myocardial perfusion and a breast cancer imaging agent.17 It is retained in the heart and breast tumors with a long half-life; an important aspect of its success, however, is its fast clearance from the non-target tissues. This fast clearance from background tissues results from hydrolysis of the methoxy group and subsequent formation of more highly charged hydrophilic species, and this hydrolysis happens at a much slower rate in the heart.18-19 Cardiotec™ is a seven coordinate neutral T c m complex and is also approved as a myocardial perfusion agent. 9 9 m Tc(CDO)(CDOH) 2 BCH 3 is a member of the BATO class of complexes. The chloride ligand is labile under physiological conditions, and the resultant aqua complex exists in equilibrium with its hydroxo analogue (7 < pK a < 7.4). This equilibrium is thought to affect its clearance from the heart: the neutral complex (CI or OH) may be washed out of the heart, and the cationic aqua species may be retained.3 These examples led to the concept of pro-drugs, complexes which have built in reaction sites that can be modified in vivo. It also challenged the notion of the necessity of thermodynamic stability/kinetic inertness in a drug: (some of) the complexes are deliberately destabilized in order to enhance the target specificity. To a chemist, a challenge in the small molecule approach is, therefore, how to introduce in vivo reactivity into a metal complex. 4 H 3 CH 2 COOC - N ^ N H - C O O C H 2 C H 3 TcO(^ / -HM-PAO) (Ceretec™) TcO(L,L-ECD) (Neurolite™) TcCI(CDO)(CDOH) 2 BMe (Cardiotec™) N III C I C R—N=C-Tc—C—N—R N * c A R ' R I R R = C H 2 C ( C H 3 ) 2 O C H 3 Tc[(2-methoxy-2-methylpropyl)isonitrile]6 (Cardiolite™) S c h e m e 1.1 Target-specific drug design attempts to control the site of drug localization beyond physical properties of the metal complex.7'9 It typically employs a bifunctional approach: a high affinity receptor ligand (or biologically active molecule, BAM) is conjugated to a radionuclide through a bifunctional chelator (BFCA), and a linker is often used to modify the pharmacokinetics of the molecule. Despite many difficulties associated with this approach, active research progressed steadily and a number of new complexes are approved or under 5 clinical investigation. The biodistribution of a target-specific radiopharmaceutical depends largely on the affinity between the receptor and its receptor ligand, but the tagged metal chelate and the linker is not completely innocent, because they contribute significantly to the overall size and charge of the whole molecule. Therefore the design of the BFCA/radionuclide fragment is very important, and it should bear easily modifiable units in order for it to adapt to any change when necessary. The synthetic challenge in the B F C A approach is bioconjugation: how to link the B A M to a coordination complex. Ternary complexes should be of interest in all three approaches. In the direct labeling approach, tagging a ternary complex to MAb's could conceivably be less disruptive to the native structure of the antibody than simply using the metal ion. Because there are fewer coordination sites to fill on the metal complex, fewer functional groups on the MAb will be disrupted by the metal binding. Ternary complexes also offer the possibility of biomodification on the metal site: substitution of one ligand would most likely affect the biochemistry of the complex. In the BFCA approach, the ternary complexes typically are stabilized by two or three complementary ligands, one of which, often the smaller ligand for easy synthesis, can be conjugated to the B A M and the remaining coordination sites can be filled by other ligand(s). They also provide an opportunity for easy modification: the coligands can be individually varied and the coordination chemistry of the complex can be altered in order to modify the chemical and physical characteristics of the molecule. The chemistry of Re is rich, and ternary complexes are ubiquitous. With suitable ligands, complexes in oxidation states from 0 to +7 have been synthesized.1'2'20-21 With a prerequisite of stability/inertness in air and in water, complexes with a combination of O, S, N, P and halide donor ligands are common, even those containing alkyl donors are often 6 seen. In most oxidation states the metal binds more avidly to S or P donors than to N or O donors. The following discussion of Re chemistry will be limited to what is relevant for this thesis. High oxidation state (+5 to +7) complexes typically contain Re=0 or Re=N cores stabilized by a combination of neutral and anionic ligands.1'2'2 0'2 1 By far the most prevalent oxidation state in rhenium chemistry (utilized in the field of nuclear medicine) is five, where [Re=0] 3 +, trans [0=Re=0] +, linear [Re20a] 4 + and [Re=N] 2 + complexes predominate. While these complexes with halides, or monodentate nitrogen/oxygen donor ligands, are not stable with respect to oxidation or hydrolysis, those with multidentate ligands often are. With only a few exceptions, in which they are anionic (mostly with halide and pseudohalide donors),2 2'2 3 the complexes are usually neutral or monocationic. Many linear tetradentate N x 0 4 - X or NXS4_X ligands have been developed to stabilize the metal centre. With N2O2 ligands, the complexes either adopt the Re20sL2 configuration, or they coordinate with a halide or other anionic ligand cis to the oxo O atom. 2 4 - 3 4 ReO(N x S4- x ) complexes, however, are predominantly five-coordinate with square pyramidal geometry around the metal centre (Scheme 1.2).35'36 They are usually very stable and these ligands have been widely used as metal chelators in B F C A - B A M conjugates. A class of ternary "3+1" complexes, in which a monodentate thiolate, bearing a receptor moiety, complements a tridentate NS2 or S3 donor set ligand, is very successful in bioconjugation chemistry.37'38 As an illustration of the kinetic difference between Tc and Re, a major difficulty facing the Tc "3+1" complexes is thought to be a fast in vivo substitution of the monothiolate by glutathione, a reaction which is very slow in the Re analogue.39'40 7 N N N > -Re X X = halide V I I v General coordination mode for a N 2 0 2 ligand to Re v c General coordination mode for a N X S 4 . X ligand to R e v S S - R "3+1" X = N R o r S - N O N T [ R e 0 2 N 4 ] + ReO(N 3 )X 2 Scheme 1.2 With neutral linear N donor ligands (or macrocyclic N4 ligands), [Re v 02N4] + coordination predominates (Scheme 1.2).41"44 These complexes are typically monocationic, and they are very stable. It has been demonstrated with ethylenediamine that the oxo O atom can be protonated ([ReO(OH)(en)2]2+ has been isolated).45'46 Tripodal amines, e.g. tacn, N(CH2py)3, coordinate well to the Re=0 core with additional anionic donors giving overall neutral or cationic complexes.4 7 - 5 5 8 Re substitution reactions are particularly slow, but the formation of M O 2 , a material that has been termed a "thermodynamic sink" in Tc chemistry, is less prevalent in Re chemistry.21 Most Re I V complexes known in the field of nuclear medicine, with the exception of Re-HEDP, are with phosphine ligands.1-9 A few Re111 complexes have been made with multidentate ligands and the metal centers were found to be "naked", or to have chloro/aqua co-ligands. For example, the Re analogue of BATO and tripodal aminephenol complexes of Re have been made. 5 6' 5 7 An important class of Re111 radiopharmaceuticals under development is Re(HYPY)xLy, where the HYPY is an oxidized form of hydrazinopyridine, a derivative of which, hydrazinonicotinic acid (HYNIC) has been easily conjugated to a B A M (Scheme 1.3).58'59 Low oxidation state Re complexes are often stabilized by (a) 71-acceptor ligand(s), in combination with neutral and anionic donor ligands. A family of Re(CO)3L3 complexes have been made and their derivatives are being investigated for radiopharmaceutical applications.60 + ReCI 3(r| 1-NNC5H4NH)(Ti2-HNNC5H4N) (ReCI 3 (HYPY) 2 ) fac- [Re(CO) 3 (H 2 0) 3 f Scheme 1.3 The redox chemistry of Re complexes is extensive. They are more easily oxidized than their Tc counterparts, and reducing perrhenate to intermediate oxidation state halo/aqua 9 complexes often requires severe conditions.21 (The term "thermodynamic sink" in aerial Re complexes could be used to describe perrhenate.) With a suitable ligand, however, the reduction becomes much easier, and reduction/complexation reactions often require milder reaction conditions and give higher yields than do substitution reactions. For example, Re(CO)3(H20)3 and a number of ReO(S2N)S and ReO(S3)S complexes can be made by mixing perrhenate with the respective ligands in the presence of a reductant.38-60 Of particular interest, perrhenate is reduced by hydrazinopyridine (HYPY) followed by concomitant complexation of the oxidized HYPY in the presence of coligands forming Re(HYPY) x L y . 5 8 ' 5 9 Quite often, when the Re-L (L = multidentate ligand) core is defined, it is stable in a number of oxidation states with the aid of different co-ligands. For example, tacn (1,4,7-triazacyclononane) complexes of Re have been isolated as [(tacn)ReV I I03]+, [(tacn)Re vOX 2]+, [(tacn)Re1(CO)3]+ and a number of Re111 and Re I V species.4 8-6 1'6 2 Thus the task of making ternary complexes in high yield in one-pot syntheses from perrhenate is not as daunting as it sounds. In practice, one more restriction obtains: in a "kit" formulation where perrhenate (10"6 -10" M) is reduced in the presence of a large excess of the ligand(s), a reductant, NaCl and water, simple stoichiometric control is not possible. Thus choosing the right combination of the ligands is crucial: the ligands must complement rather than compete with each other, and the combination of the ligands should stabilize the complex better than each of the ligands does alone. Stemming from this is the simple notion of charge balance: a combination of neutral and anionic donor ligands should fulfill the charge demand of a target complex. For example, the syntheses of "3+1" complexes explore the preference of ReO thiolate complexes to be neutral rather than cationic. This consideration alone is not sufficient 10 because competition of chloro/aqua ligands is always present. (Chloride is an excellent donor 1 RR 1 RR for Re and NaCl solution is used to elute the perrhenate in the W/ Re generator.) Also important is cr-donor, 71-donor and 7t-acceptor combinations for complexes in different oxidation states. An excellent example can be found in Tc chemistry. When pertechnetate is reduced/complexed with HYNIC-peptide in the presence of tricine, a number of unstable complexes can be detected, but if the same reaction is carried out with a triphenylphosphine derivative, only two diastereomers form (resulting from the chiral centers on the peptide backbone and the chiral Tc chelate), and they are both stable (Scheme 1.4).63 peptide C \ .NH + NaTc0 4 + P R 3 + tircine + SnCI 2 Scheme 1.4 This thesis mainly describes the synthesis and characterization of Re complexes with their radiopharmaceutical applications in mind. Specifically, we want to develop new ternary complexes that can be easily synthesized from perrhenate. Chapter 2 presents a [ReO(N202)X] system and discusses the possibility of one-pot syntheses of "4+1" complexes from perrhenate. Chapter 3 deals with the synthesis and structural features of [RevON4X2]+ complex cations, and describes a potentially general seven-coordinate motif. Chapter 4 is a 11 further study on Re chemistry with the N4 ligand in Chapter 3, N,N'-bis(2-pyridylmethyl)ethylenediamine (Ffipmen). Three Rev-pmen complexes have been made and they are related by hydrolysis. Ternary complexes with 3-hydroxy-4-pyrones have been made in one-pot syntheses from [ReCu]" via a curious dehydrogenation reaction. 12 References 1) Clarke, M. ; Podbielski, L. Coord. Chem. Rev. 1987, 78, 253. 2) Deutsch, E. ; Libson, K.; Jurisson, S.; Lindoy, L. F. Prog. Inorg. Chem. 1983, 30, 75. 3) Jurisson, S.; Desilets, C. P.; Jia, W.; Ma, D. Chem. Rev. 1993, 93, 1137. 4) Dilworth, J. R.; Parrott, S. J. Chem. Soc. Rev. 1998, 27, 43. 5) Tisato, F.; Refosco, F.; Bandoli, G. Coord. Chem. Rev. 1994,135, 235. 6) Volkert, W. A.; Hoffman, T. J. Chem. Rev. 1999, 99, 2269. 7) Jurisson, S. S.; Lydon, J. D. Chem. Rev. 1999, 99, 2205. 8) Heeg, M . J.; Jurisson, S. S. Acc. Chem. Res. 1999, 32, 1053. 9) Liu, S.; Edwards, D. S. Chem. Rev. 1999, 99, 2235. 10) Deklerk, J. M . H.; Vanhetschip, A. D.; Zonnenberg, B. A.; Vandijk, A.; Quirijnen, J. M . S. P.; Blijham, G. H.; Vanrijk, P. P. J. Nucl. Med. 1996, 37, 244. 11) Griffithes, G. L.; Goldenberg, D. M . ; Jones, A. L.; Hansen, H. J. Bioconjugate Chem. 1992, 5,91. 12) Breitz, H. B.; Fisher, D. R.; Wessels, B. W. J. Nucl. Med. 1998, 39, 1746. 13) Juweid, M . ; Sharkey, R. M . ; Swayne, L. C ; Griffiths, G., L.; Dunn, R.; Goldenberg, D. M . J. Nucl. Med. 1988,19, 34. 14) Jurisson, S.; Schlemper, E. O.; Troutner, D. E.; Canning, L. R.; Nowotnik, D. P.; Neirinckx, R. D. Inorg. Chem. 1986, 25, 543. 15) Neirnckx, R. D.; Canning, L. R.; Piper, I. M . ; Nowotnic, D. P.; Pickett, R. D.; Holmes, R. A.; Volkert, W. A.; Forster, A. M . ; Weisner, P. S.; Marriott, J. A.; Chaplin, S. B. J. Nucl. Med. 1987, 2,191. 13 16) Hung, J. C ; Corlija, M . ; Volkert, W. A.; Holmes, R. A. J. Nucl. Med. 1988, 1568. 17) Abrams, M . J.; Davison, A.; Jones, A. G.; Costello, C. E.; Pang, H. Inorg. Chem. 1983, 22, 2798. 18) Kronauge, T. F.; Leon, A. S.; Verdera, E. S.; Baiter, H. S.; Leon, E. T.; Mut, F.; Oliveira, M . C.; Garcia, F. A.; Holman, B. L.; Davison, A.; Jones, A. G. J. Nucl. Med. 1991, 33, 1949. 19) Kronauge, J. F.; Chin, M . L.; Cone, J. S.; Davison, A.; Holman, B. L.; Jones, A. G.; Piwnica-Worms, D. Nucl. Med. Biol. 1992,19, 141. 20) Rouschias, G. Chem. Rev. 1974, 74, 531. 21) Connor, K. A.; Walton, R. Comprehensive Coordination Chemistry; Wilkinson, G., Gillard, R. D. and McCleverty, J. A., Ed.; Pergamon Press: Oxford, 1987; Vol. 4. 22) Dilworth, J. R.; Hu, J.; Miller, J. R.; Hughes, D. L.; Zubieta, J. A.; Chen, Q. J. Chem. Soc. Dalton Trans. 1995, 3153. 23) Cotton, F. A.; Lippard, S. J. Inorg. Chem. 1966, 5, 9. 24) Banbery, H. J.; McQuillan, F.; Hamor, T. A.; Jones, C. J.; McCleverty, J. A. J. Chem. Soc. Dalton Trans. 1989, 1405. 25) Bandoli, G.; Nicolini, M. ; Mazzi, U.; Refosco, F. J. Chem. Soc, Dalton Trans. 1984, 2505. 26) Herrmann, W. A.; Rauch, M . U.; Arms, R. J. Inorg. Chem. 1996, 35,1988. 27) Jurisson, S.; Dancey, K. P.; McPartlin, M. ; Tasker, P. A.; Deutsch, E. Inorg. Chem. 1984, 23, 4743. 28) Jurisson, S.; Lindoy, L. F.; Dancey, K. P.; McPartlin, M . ; Tasker, P. A.; Uppal, D. K.; Deutsch, E. Inorg. Chem. 1984, 23, 227. 29) Middleton, A. R.; Masters, A. F.; Wilkinson, G. J. Chem. Soc, Dalton Trans. 1979, 542. 14 30) Refosco, F.; Tisato, F.; Mazzi, U.; Bandoli, G.; Nicolini, M . J. Chem. Soc., Dalton Trans. 1988,611. 31) Gilli, G.; Sacerdoti, M . ; Bertolasi, V.; Rossi, R. Acta Cryst. 1982, B38, 100. 32) Tisato, F.; Refosco, F.; Mazzi, U.; Bandoli, G.; Dolmella, A. Inorg. Chirh. Acta 1989, 164, 127. 33) Tisato, F.; Refosco, F.; Mazzi, U.; Bandoli, G.; Nicolini, M . Inorg. Chim. Acta 1991, 189, 97. 34) van Bommel, K. J. C.; Verboom, W.; Kooijman, H.; Spek, A. L.; Reinhoudt, D. N. Inorg. Chem. 1998,57,4197. 35) Chi, D. Y.; O'Neil, J. P.; Anderson, C. J.; Welch, M . J.; Katzenellenbogen, J. A. J. Med. Chem. 1994, 57, 928. 36) Hansen, L.; Lipowska, M . ; Melendez, E.; Xu, X.; Hirota, S.; Taylor, A. T.; Marzilli, L. G. Inorg. Chem. 1999, 38, 5351 and references therein. 37) Meegalla, S.; Plossl, K.; Kung, M.-P.; Stevenson, D. A.; Liable-Sands, L. M . ; Rheingold, A. I.; Kung, H. F. J. Am. Chem. Soc. 1995,117, 1137. 38) Papadopoulos, M . S.; Pirmettis, I. C ; Pelecanou, M . ; Raptopoulou, C. P.; Terzis, A.; Stassinopoulou, C. I.; Chiotellis, E. Inorg. Chem. 1996, 35, 7377, and references therein. 39) Nock, B. A.; Maina, T.; Yannoukakos, D.; Pirmettis, I. C ; Papadopoulos, M . S.; Chiotellis, E. J. Med. Chem. 1999,1066, and references therein. 40) Syhre, R.; Seifert, S.; Spies, H.; Gupta, B.; Johannsen, B. Eur. J. Nucl. Med. 1998, 25, 793. 41) Beard, J. H.; Casey, J.; Murmann, K. Inorg. Chem. 1965, 4, 797. 42) Belanger, S.; Fortin, S.; Beauchamp, A., L. Can. J. Chem. 1997, 75, 37. 15 43) Lock, C. J. L.; Turner, G. Acta Crystallogr. Sect. B 1978,34, 923. 44) Masood, A.; Hodgson, D. J. Inorg. Chem. 1994, 33, 2488. 45) Technetium, Rhenium and Other Metals in Chemistry and Nuclear Medicine; S. G. Editorali: Padova, 1999; Vol. 5. 46) Fergusson, J. E. Coord. Chem. Rev. 1966, 459. 47) Al-Shihri, A. S. M . ; Dilworth, J. R.; Howe, S. D.; Silver, J.; Thompson, R. T. Polyhedron 1993,12, 2297. 48) Bohm, G.; Wieghardt, K.; Nuber, B.; Weiss, J. Inorg. Chem. 1991, 30, 3464. 49) Brown, S. N.; Mayer, J. M . Inorg. Chem. 1992, 31, 4091. 50) Brown, S. N.; Mayer, J. M . J. Am. Chem. Soc. 1996,118, 12119. 51) Paulo, A.; Domingos, A.; Marcalo, J.; de Matos, A. P.; Santos, I. Inorg. Chem. 1995, 34, 2113. 52) Paulo, A.; Domingos, A.; Santos, I. Inorg. Chem. 1996, 35, 1798. 53) Sugimoto, H.; Sasaki, Y. Chem. Lett. 1997, 541. 54) Sugimoto, H.; Kamei, M . ; Umakoshi, K.; Sasaki, Y.; Suzuki, M . Inorg. Chem. 1996, 35, 7082. 55) Takahira, T.; Umakoshi, K.; Sasaki, Y. Acta Crystallogr., Sect. C1994, 50, 1870. 56) Jurisson, S.; Halihan, M . M . ; Lydon, J. D.; Barnes, C. L.; Nowotnik, D. P.; Nunn, A. D. Inorg. Chem. 1998, 37, 1922, and references therein. 57) Wong, E. S. Y. Ph. D. Thesis; University of British Columbia: Vancouver, 1996. 58) Hirsch-Kuchma, M . ; Nicholson, T.; Davison, A.; Davis, W. M. ; Jones, A. G. Inorg. Chem. 1997, 36, 3237. 16 59) Rose, D. J.; Maresca, K. P.; Nicholson, T.; Davison, A.; Jones, A. G.; Babich, J.; Fischman, A.; Graham, W.; DeBord, R. D.; Zubieta, J. Inorg. Chem. 1998, 37, 2701. 60) Alberto, R.; Schibli, R.; Egli, A.; Schubiger, A. P. J. Am. Chem. Soc. 1998,120, 7987. 61) Romao, C. C.; Kuhn, F. E.; Herrmann, W. A. Chem. Rev. 1997, 8, 3197. 62) Bohm, G.; Wieghardt, K.; Nuber, B.; Weiss, J. Angew. Chem. Int. Ed. Engl. 1990, 29, 787. 63) Edwards, D. S.; Liu, S.; Barrett, J. A.; Harris, A. R.; Looby, R. J.; Ziegler, M . C.; Heminway, S. J.; Caroll, T. R. Bioconjugate Chem. 1997, 8, 146. 17 Chapter 2. [ReO(N202)X] Complexes: 4+1? 2.1 Introduction Following up on a study of tripodal amine phosphinate ligands and their complexes with lanthanides and group 13 metals, we made FLmpmeCh, an N2O2 donor set amine phosphinate ligand for Re v=0 or Tc v =0 complexation (Scheme 2.1).1'2 Volumes of data exist on Re/Tc=0 complexes with tetradentate N 2 0 2 ligands, but most studies have been done with Schiff base or amine phenolate/enolate systems.3-10 The complexes are usually hydrolytically sensitive, attributed sometimes to the instability of Schiff base imine linkages. Also, the O donors in these ligands are mostly phenolates or enolates that binds well under basic conditions. To our knowledge, no N2O2 ligands with carboxylate, phosphonate, or phosphinate O donors, i.e. those that bind under neutral or acidic conditions, have been successfully complexed to this M v = 0 core. Given the hydrolytic instability of Tc and ease of oxidation of Re at high pH (where the phenolates and enolates binds best), the incorporation of phosphinates should introduce a bias for the ligand to coordinate in low pH reaction conditions, favourable to both the reduction of the metal ion and the stability of the complex. 2+ H 3 C X / \ / C H 3 2cr o H4ppmeCl2 Scheme 2.1 18 We are also interested in ternary complexes. Dianionic N2O2 ligands usually complex M v = 0 with an additional alcoholate, halide or bridging oxo ligand, forming overall neutral complexes.3"10 In light of a series of successful "3+1" complexes, in which a monodentate thiolate, bearing a receptor moiety, complements a tridentate N S 2 or S3 donor set ligand, 1 1 - 1 5 we think these N2O2 donors have the potential of forming "4+1" complexes with monodentate thiolates. 19 2.2 Experimental Materials. N,N'-Dimethylethylenediamine, phenylphosphinic acid, 30% formaldehyde, 37% HC1, potassium thiocyanate, benzene thiol, 4-methoxythiophenol, lipophilic Sephadex LH-20, and tetrabutylammonium bromide were obtained from Aldrich, Alfa, Fisher or Sigma, and used without further purification. [NH4][Re04] was received as a gift from Johnson-Matthey. [(C4H9)4N][ReOBr4]16 was synthesized by literature methods and FLmpmeCb was first prepared by Dr. Mark P. Lowe in our group. Instrumentation. IR (infrared) spectra were recorded as KBr discs in the range 4000-500 cm"1 on a Mattson Galaxy Series 5000 FTIR spectrophotometer. Mass spectra (Cs+, LSIMS (Liquid Secondary Ion Mass Spectrometry)) were obtained on a Kratos Concept IIH32Q with thioglycerol as the matrix. C, H, and N analyses were performed by Mr. Peter Borda in this department. All 'H NMR spectra were recorded on a Bruker A C -31 200E spectrometer at 200 MHz, and are reported as 5 in ppm downfield from TMS. All P {'H} NMR spectra were recorded on the same instrument at 81 MHz, and are reported as 5 in ppm downfield from external phosphoric acid. H P L C Experiments were performed with Waters 501 HPLC pumps, a Waters U6K injector, and a Waters Lambda-Max Model 481 L C spectrophotometer, all interfaced to a PC. A Hamilton PRP-1 reverse phase column (15 cm) was used; all runs were isocratic using 50/50 acetonitrile/water as the eluent. Gradient experiments were performed but no extra peaks were found. After some of the experiments, acetonitrile or methanol were used to 20 elute the column but no extra species was detected in 40 min. The experiments were performed with the detector set at 350, 475, 600 and 698 nm for each sample. H4ppmeCl2. Phenylphosphinic acid (3.5 g, 25 mmol) and N,N'-dimethylethylenediamine (1.0 g, 11 mmol) were dissolved in 6 M HC1 (40 mL). The solution was heated to reflux, and 37% formaldehyde (4.9 g, 57 mmol) was added dropwise over 20 min. The solution was refluxed for 4.5 h and then concentrated to yield a white solid, which was dissolved in methanol (200 mL), and diethyl ether (100 mL) was added to precipitate the white product; the yield was quantitative. Anal. Calcd (Found): for C18H28CI2N2O4P2: C, 46.07 (46.27); H, 6.01 (6.24); N, 5.97(5.86). Mass spectrum (LSIMS): m/z 397 ([H3ppme]+). 3 1P{ 1H} N M R (D 20) 8: 19 (s). 'H N M R (D 20) 5: 7.8-7.4 (10 H, 2 sets of multiplets, PC 6 H 5 ) , 3.7 (4 H, s, ethylene), 3.5 (4 H, d, NCH 2 P), 2.9 (6 H, s, NCH 3 ) . [ReO(ppme)X] X = Bro.3/00.7. To a stirred ethanolic solution (10 mL) of FLmpmeCfi (62 mg, 0.13 mmol) was added [(C4H9)4N][ReOBr4] (100 mg, 0.13 mmol) dissolved in 10 mL ethanol. A gray precipitate formed immediately but redissolved when triethylamine (63 mg, 0.63 mmol) was added drop-wise. The blue solution was refluxed for 2 h and was concentrated. The concentrated solution was loaded onto a Sephedex LH-20 column and eluted with ethanol. The blue fraction was collected, concentrated and dried under vacuum to yield 59 mg (~ 70 %). Mass spectrum (LSIMS): m/z 699 ([ReO(ppme)Br+Na]+), 677 ([ReO(ppme)Br+H]+), 655 ([ReO(ppme)Cl+Na]+), 633 ([ReO(ppme)Cl+H]+), 597 ([ReO(ppme)]+). 3l?{lU} NMR (CD 3CN) 5: 84 (two peaks, AS = 0.3), 38 (two peaks, A8 = 0.5). A microcrystalline material precipitated when diethyl ether diffused into an ethanolic 21 solution of the product. Single crystals precipitated from a C D 3 C N solution of the crystalline material. [ReO(ppme)Cl]«H 20. To a hot methanolic solution (10 mL) of H 4ppmeCl 2 (142 mg, 0.300 mmol) and [NH4][Re04] (157 mg, 0.586 mmol), triphenylphosphine (90 mg, 0.34 mmol) was added. The mixture was heated for 20 h at 70°C to yield a blue solution and a small amount of precipitate. The solution was decanted, concentrated, and dried under vacuum. The resultant blue residue was suspended in acetone and the suspension was filtered. Diethyl ether was added liberally to the filtrate to precipitate the blue product, which was filtered out and dried to yield 180 mg (91.6%). A crystalline material formed on slow evaporation of an acetone/water solution of the product. Anal. Calcd (Found): for CigH26ClN206P2Re: C, 33.26 (33.03); H, 4.03 (3.98); N, 4.31 (4.03). IR (cm"1, KBr disk): 1206(vPO), 1199 (v r o), 1124 (v P 0), 949, 931. Mass spectrum (LSIMS): m/z 633 ([ReO(ppme)Cl+H]+), 597 ([ReO(ppme)]+). ^P^H} NMR (acetone-J6): 84, 38. HPLC: peak maximum at 1.2 min. [ReO(ppme)NCS]CH3OH. To a methanolic solution (3 mL) of [ReO(ppme)Cl]»H 2 0 (65 mg, 0.10 mmol) was added KSCN (11 mg, 0.11 mmol) resulting in immediate precipitate of a white salt. The supernatant was decanted, and was allowed to stand at room temperature for two days upon which time more white precipitate formed and was subsequently filtered out. The filtrate was concentrated and dried. Anal. Calcd (Found): for C20H28N3O6P2ReS: C, 34.98 (35.30); H, 4.11 (4.00); N, 6.12 (5.73). IR (cm"1, KBr disk): 2060 (vC N), 1206 (v P 0), 1123 (vp0), 945, 923. Mass spectrum (LSIMS): m/z 656 ([M+H]+), 597 ([M-NCS]+). ^P^H} (acetone-<i6) N M R 8: 79, 38. HPLC: peak maximum at 2.0 min. 22 [ReO(ppme)(SC6H5)]. Method A. [ReO(ppme)Cl]*H20 (23 mg, 0.035 mmol) was dissolved in 1 mL acetone-cfc and KSC6H5 (8 mg, 0.05 mmol) was added. The mixture was heated at 70°C for 40 h, during which time a white precipitate slowly formed and the solution turned brown. The supernatant was decanted and the 3 1 P NMR spectrum recorded. Method B. ReO(ppme)Cl»H 2 0 (63 mg, 0.10 mmol) and [H3ppme][ReO(SC6H5)4] (same preparation as [H3ppme][ReO(SC6H5OCH3)4]-2H20, 20 mg, 0.019 mmol) were dissolved in 2 mL C D 3 C N with a few drops of D 2 0 . The mixture was heated at 70°C for 40 h, upon which time [ReO(ppme)(SC6H5)], [ReO(ppme)Cl] and [H3ppme][ReO(SC6H5)4] were seen in the 3 1 P and lH N M R spectra. More [H3ppme][ReO(SC6H5)4] (a total of 48 mg, 0.046 mmol) was added in two portions, each followed by 20 h heating. More D 2 0 was added, resulting in the formation of a very dark brown oily precipitate which was allowed to settle. The supernatant was decanted and more D 2 0 was added in order to precipitate the brown product. The solvent was further reduced by rotary evaporation and was decanted off. The brown product was dried under vacuum. Anal. Calcd (Found): for C 2 4H 29N 205P 2ReS: C, 40.85 (40.71); H, 4.14 (4.19); N, 3.97 (4.20). IR (cm"1, KBr disk): 1200 (v P 0), 1126 (vP O), 952, 930. Mass spectrum (LSIMS): m/z 1413 ([2M+H]+), 707 ([M+H]+), 597 ([M-C 6H 5S] +). 3 1P{'H} NMR (CD 3CN) 8: 74, 34. HPLC: peak maximum at 2.3 min. [H3ppme][ReO(SC6H40CH3)4]-2H20. To a solution of H 4ppmeCl 2 (94 mg, 0.20 mmol) and [NH4][Re04] (54 mg, 0.20 mmol) in 5 mL methanol was added H S C 6 H 4 O C H 3 (180 mg, 1.29 mmol). The solution immediately turned dark brown and a red/brown microcrystalline material precipitated within 20 min. After two hours, the crystals were filtered out, washed 23 with methanol and dried under vacuum to yield 199 mg (86%). Anal. Calcd (Found): for C46H59N2OiiP2ReS4: C, 46.34 (46.59); H, 4.99 (4.85); N, 2.35(2.35). IR (cm - 1, KBr disk): 1589,1174 (vP O), 959 (v R e 0). 3 1P{ !H} NMR (DMSO-</6) 5: 25. *H NMR (DMSO-40 °: 7.9 -7.4 (10 H, 2 sets of multiplets, PC 6 H 5 ) , 7.4 and 6.9 (8 H each, 2 sets of doublets, SC 6 H 4 0) , 3.8 (12 H, s, OCH 3 ) , 3.6 - 3.4 (8 H, overlapped methylene H), 2.7 (6 H, s, NCH 3 ) . [ReO(ppme)(SC 6H 4OCH 3)]. Same as method B for [ReO(ppme)(SC6H5)], and [H3ppme][ReO(SC6H4OCH3)4] (a total of 54 mg, 0.046 mmol, in 4 portions) was used. Anal. Calcd (Found): for C 2 5 H3iN 2 0 6 P 2 ReS: C, 40.81 (40.91); H, 4.25 (4.49); N, 3.81(3.50). IR (cm"1, KBr disk): 1198 (vP O), 1122 (v P 0), 950, 930. Mass spectrum (LSIMS): m/z 1473 ([2M+H]+), 737 ([M+H]+), 597 ([M-C 7H 7OS] +). 31P{'H} NMR (acetone-^) 8: 80, 38. HPLC: peak maximum at 2.3 min. 24 2.3 Results and Discussion Under anhydrous conditions and in the presence of a base, FLmpmeCb reacts with [(C4H9)4N][ReOBr4J to form [ReO(ppme)X] (X = Br/CI), a mixture of halide complexes which are difficult to separate. In contrast, ammonium perrhenate can be reduced with triphenylphosphine in the presence of the ligand to form [ReO(ppme)Cl]-H20 in a vial, and in high yield. E A of the blue crystalline material agrees with its formulation, LSIMS shows [M+H]+ and [M-C1]+ peaks, the IR spectrum shows prominent PO peaks as well as two possible VRe=o at 931 and 949 cm"1 (identification of the Re=0 peak was not attempted). The 3 1 P N M R spectrum exhibits two resonances at 84 and 38 ppm, indicating a drastic difference in the coordinating phosphinates: one is cis and the other trans to the oxo O atom. Correspondingly, the 'H NMR spectrum shows that the molecule is asymmetric and that the two sides of the ligand are in markedly different chemical environments. The reduction/complexation reaction is interesting. In a 1:1:1 methanolic mixture of [NH4][Re04], triphenylphosphine, and FLjppmeCb, [ReO(ppme)Cl] formed almost exclusively ( 3 1P NMR); triphenylphosphine oxide and a very small amount of triphenylphosphine-related complexes were also formed. The presence of excess triphenylphosphine impeded the substitution reaction because ReO(P(C6Hs)3)2Cl3 was formed and remained intact for days in the reaction mixture. Similarly, ReO(P(C6H-5)3)2Cl3 did not react with ppme in a simple substitution fashion. Single crystals of [ReO(ppme)X] (X = Br/CI) precipitated from a C D 3 C N solution of the complex mixture and the X-ray crystallographic analysis confirmed the structure proposed based on N M R results (Figure 2.1 and Table 2.1). The Re center is coordinated by 25 the N2O2 ligand, an oxo O and a halide atom, forming a distorted octahedron. The halide atom is coordinated cis to the oxo atom; and the two N-methyls are syn to the oxo O. The cis coordination of the halide atom to the oxo atom is common, and to our knowledge, it is the only mode of coordination in ReO(N202)Cl complexes.9'10 The distortion, where the ligand coordinating sites squeeze towards each other and bend away from the oxo O, is also typical of a ReO(N2C»2)X complex. The Re=0 bond is 1.666 A, and is at the short end of the reported values; the Re-N, Re-0 and Re-Cl bond lengths are within the expected range. C(8) Figure 2.1. ORTEP Drawing of [ReO(ppme)Br0.3Clo.7]- Thermal Ellipsoids for the Non-Hydrogen Atoms are Drawn at the 33% Probability Level. 26 Table 2.1. Selected Bond Lengths (A) and Bond Angles (deg) in [ReO(ppme)Br0.3Clo.7]. Re-0(5) 1.666(6) Re-N(l) 2.187(7) Re-N(2) 2.139(6) Re-0(3) 1.995(6) Re-O(l) 2.031(5) Re-Br/Cl 2.48(1)/2.38(1) Br/Cl-Re-0(5) 96.0(4)/96.2(4) N(l)-Re-0(5) 88.8(3) N(2)-Re-0(5) 92.1(3) 0(3)-Re-0(5) 109.1(3) 0(l)-Re-0(5) 166.5(3) N(l)-Re-N(2) 83.3(3) N(l)-Re-0(3) 159.7(2) N(l)-Re-0(1) 78.3(2) N(l)-Re-Br/Cl 97.2(5)/98.9(4) N(2)-Re-0(1) 82.6(2) N(2)-Re-0(3) 86.6(2) Unlike other ReO-(N202) complexes, where coordination modes vary and a few diasteromers persist, one [ReO(ppme)X] diastereomer forms preferentially. Also, the halide position is not labile: simple solvolysis of the halide in CD3OD or D2O was not seen even over a day or two. (The complex, however, decomposes in a heated CD3OD/D2O mixture over a few weeks.) These features suggest an excellent opportunity for conjugating the [ReO(ppme)]+ unit to a monodentate ligand through a halide substitution reaction. Following 27 the success of the aforementioned series of "3+1" complexes,11"15 we probed a "4+1" system. [ReO(ppme)Cl]H20 reacts with KSCN or K S C 6 H 5 to form [ReO(ppme)(NCS)] or [ReO(ppme)(SC6H5)]. [ReO(ppme)(SC6H4OCH3)] was also made but from a different route (vide infra). The products were characterized by LSIMS, IR, and the 3 1 P N M R spectra, and their purity was checked by E A and HPLC. The IR spectra of all the [ReO(ppme)X] complexes are very similar. The diagnostic vpo and VRe=o stretches are evident in every complex, but vary slightly in frequency. Additional ligand peaks are seen for the respective coligand. The NCS" is postulated to be N-bound in [ReO(ppme)NCS] based on analogous V C N values.17 (The C N stretching frequencies in N4?onded complexes are near 2050 cm"1, and that in S-bonded complexes are near 2100 cm"1, but C N stretching alone is not definitive evidence in such determination.17) The (+)LSIMS of all the ReO(ppme)X complexes invariably exhibit prominent [M+H]+ and [M-X] + peaks, the 3 1 P NMR spectra show two peaks at -80 and -35 ppm, and the *H NMR spectra display similar resonances resulting from the ppme backbone. These features demonstrate that the complexes share the same structure with respect to [ReO(ppme)]+ and only differ in the X coordination. These complexes also have similar solubility in water, organic and mixed solvents. Related to that, HPLC experiments show that these complexes are similarly retained on a reverse phase column. 3 1 P NMR experiments following these reactions suggested that the substitutions were reasonably simple, and that only the products and the free ligand (ppme2") were found in the mixture. The free ligand formation can be explained by decomposition of the complex and/or failure of ppme2" to compete with the thiol. This demetallation of ppme2" is ominous 28 for a one-pot synthesis of the ternary complexes. In a pharmacological "kit" formulation where perrhenate (or pertechnetate) is reduced with both of the ligands in excess, formation of ternary complexes requires the ligands to complement, rather than to compete with, each other. An attempted one-pot synthesis of the ternary complex [ReO(ppme)SR] resulted in the formation of [H3ppme][ReO(SR)4] ("SR = "SC 6H 4OCH 3). [NH4][Re04] reacts rapidly with excess HSR in the presence of H4ppmeCl in methanol, and a red crystalline material precipitated in high yields. The immediate questions raised are whether the formation of it is driven by its insoluble nature and whether it can be converted to the desired ternary complex. [H3ppme][ReO(SR)4] was dissolved in C D 3 C N and heated or left in the open air for a few days but it did not convert to the "4+1" complex, even after addition of KOH. Thus this complex is very stable. Aside from the minor misfortune that ppme2" does not coordinate to the metal, this material is interesting. [ReO(SR)4]" complexes have been studied extensively, but to our knowledge, one-pot syntheses from perrhenate have not been reported.18'19 Recently, thiols conjugated to a biologically active molecule (BAM) have been fixed onto solid supports and then reacted with Tc precursors to form Tc-thiolate-BAM conjugates in solution, so that no free B A M was actually present in solution.20'21 This approach reduced competition of free B A M binding to the receptor. In light of this development and because [H3ppme][ReO(SR)4] has little solubility in water or alcohol, the latter could be used as solid thiol sources in forming ternary complexes. Reactions of [ReO(ppme)Cl]-H20 with [H3ppme][ReO(SR)4], however, are slow. Biphasic reaction of solid [H3ppme][ReO(SR)4] and [ReO(ppme)Cl]-H20 in methanol produces little [ReO(ppme)(SR)] even after two days at 70°C; more so, when both starting 29 materials were dissolved in acetonitrile, the formation of [ReO(ppme)(SR)] took days of heating. When the reaction was monitored by P NMR, co-existence of [ReO(SR)4]", [ReO(ppme)(SR)] and [ReO(ppme)Cl] was evident, and the ratio of the species changed when more [ReO(SR)4]" was added. Thus [ReO(SR)4]" and [ReO(ppme)Cl] species exist in equilibrium with [ReO(ppme)SR] and some side product, and the equilibrium does not lie heavily toward the formation of the ternary thiolate complexes. This equilibrium precludes the actual use of "4+1" complexes in nuclear medicine. 2.4 C o n c l u s i o n s [ReO(ppme)Cl]-H20 has been synthesized and a [ReO(ppme)X] (X = BrojClo?) sample has been structurally characterized. Substitution of CI* with thiocyanate and aromatic thiolates was successful, but the one-pot synthesis of the respective ternary thiolate complexes was not. Reduction/complexation reaction of a thiol, H2ppmeCl4, and [NH4][ReC>4] resulted in the formation of [H3ppme][ReO(SR)4], the reaction of which with [ReO(ppme)Cl]-H20 does not lead to [ReO(ppme)SR] in high yields. 30 2.5 References: 1) Lowe, M . P.; Rettig, S. J.; Orvig, C. J. Am. Chem. Soc. 1996,118, 10446. 2) Lowe, M . P.; Caravan, P.; Rettig, S. J.; Orvig, C. Inorg. Chem. 1998, 37, 1637. 3) Middleton, A. R.; Masters, A. F.; Wilkinson, G. J. Chem. Soc, Dalton Trans. 1979, 542. 4) Jurisson, S.; Lindoy, L. F.; Dancey, K. P.; McPartlin, M. ; Tasker, P. A.; Uppal, D. K.; Deutsch, E. Inorg. Chem. 1984, 23, 227. 5) Bandoli, G.; Nicolini, M . ; Mazzi, U.; Refosco, F. J. Chem. Soc, Dalton Trans. 1984, 2505. 6) Banbery, H. J.; McQuillan, F.; Hamor, T. A.; Jones, C. J.; McCleverty, J. A. J. Chem. Soc, Dalton Trans. 1989, 1405. 7) Tisato, F.; Refosco, F.; Bandoli, G. Coord. Chem. Rev. 1994,135, 235. 8) Luo, H.; Liu, S.; Rettig, S. J.; Orvig, C. Can J. Chem. 1995, 73, 2272. 9) Herrmann, W. A.; Rauch, M . U.; Artus, R. J. Inorg. Chem. 1996, 35, 1988. 10) van Bommel, K. J. C ; Verboom, W.; Kooijman, H.; Spek, A. L.; Reinhoudt, D. N. Inorg. Chem. 1998,37,4197. 11) Jurisson, S. S.; Lydon, J. D. Chem. Rev. 1999, 99, 2205. 12) Papadopoulos, M . S.; Pirmettis, I. C ; Pelecanou, M . ; Raptopoulou, C. P.; Terzis, A.; Stassinopoulou, C. I.; Chiotellis, E. Inorg. Chem. 1996, 35,1YI1, and references therein. 13) Nock, B. A.; Maina, T.; Yannoukakos, D.; Pirmettis, I. C ; Papadopoulos, M . S.; Chiotellis, E. J. Med. Chem. 1999, 1066, and references therein. 14) Syhre, R.; Seifert, S.; Spies, H.; Gupta, B.; Johannsen, B. Eur. J. Nucl. Med. 1998, 25, 793. 15) Meegalla, S.; Plossl, K.; Kung, M.-P.; Stevenson, D. A.; Liable-Sands, L. M . ; Rheingold, 31 A. I.; Kung, H. F. J. Am. Chem. Soc. 1995,777, 1137. 16) Cotton, F. A.; Lippard, S. J. Inorg. Chem. 1966, 5, 9. 17) Nakamoto, K. Infrared and Raman Spectra of Inorganic Coordination Compounds; John Wiley & Sons: New York, 1986. 18) McDonell, A. C ; Hambley, T. W.; Snow, M . R.; Wedd, A. G. Aust. J. Chem. 1983, 36, 253. 19) Dilworth, J. R.; Hu, J.; Miller, J. R.; Hughes, D. L.; Zubieta, J. A.; Chen, Q. J. Chem. Soc. Dalton Trans. 1995, 3153. 20) Pollak, A.; Roe, D. G.; Pollock, C. M . ; Lu, L. F. L.; Thornback, J. R. J. Am. Chem. Soc. 1999,727,11593. 21) Dunn-Dufault, R.; Pollak, A.; Thornback, J. R.; Ballinger, J. R. Bioconjugate Chem. 1999,10, 832. 32 Chapter 3. Seven-Coordinate [Re vON 4X 2]+ Complexes (X = O and CI)* 3.1 I n t r o d u c t i o n Re v=0 complexes are almost exclusively five-coordinate (square pyramidal geometry) or six-coordinate (octahedral geometry).2'3 To our knowledge, there is only one seven-coordinate Re=0 complex, [ReO(qtp)(OCH3)2]C104, which has a distorted pentagonal bipyramidal geometry in which the N4 donor ligand qtp forms a five-membered basal plane with the oxo O, and has two methoxides occupying axial positions.(Scheme 3.1)4 Numerous Re complexes containing pyridyl or polypyridyl ligands have been synthesized but this geometry remained unique.5"7 Previously, our group has prepared various Re(V) and Tc(V) complexes with ligands containing phosphorus, neutral oxygen and anionic oxygen donor atoms.8"11 As a follow up to those studies, seven-coordinate ReO complexes, sharing similar structural features to those in [ReO(qtp)(OCH3)2]C104, have been prepared with aminopyridyl (Ffipmen), and aminopyridylcarboxyl (F^bped), ligands (scheme 3.1). Also included in the discussion of this chapter is another ReO complex with an aminopyridylphenolate ligand, tbbbpen, also prepared in our group.12 Together these complexes represent a unique structural motif in Re=0 chemistry, one upon which I elaborate herein in a new series of seven coordinate Re=0 monocationic species which is, unexpectedly, quite general with the appropriate choice of multidentate ligands. * This work was initiated by J. Pierreroy in our group.1 33 = \ H 3 C Q ^ [ReO(qtp)(OCH3)2]+ qtp = 2,2':6',2":6",2"'-quaterpyridine HOOC COOH H2bped 6 / \ NH HN H2pmen H2bbpen Scheme 3.1 34 3.2 Experimental Materials and Instrumentation. See Chapter 2 for details. Sodium borohydride, ethylenediamine, 2-picolylchloride hydrochloride, and sodium hydroxide were obtained from Aldrich or Alfa and used without further purification. N,N'-Bis(2-pyridylmethyl)ethylenediamine-N,N'-diacetic acid dihydrochloride (H2bped-2HC1)13 and N,N'-bis(2-pyridylmethyl)ethylenediamine tetrahydrochloride (H2pmen-4HC1)1>14 were prepared as previously described. [ReOCl2(H2pmen)]Cl-CH3OH. [NH4][Re04] (52 mg, 0.20 mmol) and H2pmen-4HC1 (78 mg, 0.20 mmol) were suspended in methanol (10 mL) with heating. Triphenylphosphine (52 mg, 0.20 mmol) was added to the suspension and the mixture was heated at 70°C for 1 h resulting in a brown solution. Diethyl ether (-15 mL) was added until the solution turned slightly turbid and the mixture was kept at -10°C over a few days, over which time yellow crystals precipitated. Anal. Calcd (found) for [ReOCi2(H2pmen)]Cl-CH30H (Ci 5H22Cl3N 40 2Re): C, 30.91 (30.86); H, 3.80 (3.75); N, 9.61 (9.67). Mass spectrum (LSIMS): m/z = 515 ([ReOCl2(H2pmen)]+), 443 ([ReO(Ci 4Hi 6N 4)]+). IR (cm"1, KBr disk): 927 (vRe=o). The crystals were dissolved in methanol with a drop of concentrated HC1; diethyl ether was allowed to diffuse into the mixture and single crystals of [ReOCl2(H2pmen)]Cl-2CH3OH precipitated in a day or two. (More detailed chemistry of Re species with the Fbpmen ligand will be discussed in Chapter 4.) 35 [ReO(bped)][Re04] H 2 0 . H2bped-2HC1 (20 mg, 0.035 mmol) and [NH4][Re04] (11 mg, 0.039 mmol) were dissolved in 2 mL ethanol. Triphenylphosphine (13 mg, 0.050 mmol) was added and a gray precipitate was produced. The mixture was heated for 30 minutes and another portion of [NH4][Re04] (14 mg, 0.052 mmol) was dissolved with heating. The suspension was cooled to room temperature. The supernatant was decanted off and the gray solid was washed with ethanol and dried under vacuum. Anal. Calcd (found) for [ReO(bped)] [Re0 4 ]«H 2 0 (Ci 8 H 2 2 N 4 Oi 0 Re 2 ) : C, 26.15 (26.29); H, 2.68 (2.77); N, 6.78 (6.41). IR (cm"1, KBr disk): 920 (vRe=0). Mass spectrum (LSIMS): m/z = 559 ([ReO(bped)]+), 443 ([ReO(Ci4Hi6N4)]+). Single crystals suitable for X-ray diffraction studies were grown from a methanol solution as [ReO(bped)][Re04]-CH3OH. 36 3 . 3 Results and Discussion Both [ReO(bped)][Re04]-H20 and [ReOCl 2(H 2pmen)]ClCH 3OH can be synthesized by reducing [NH4] [Re04] with triphenylphosphine in the presence of the respective ligand. This reduction method is general (see also the synthesis of [ReO(ppme)Cl] in Chapter 1). It requires an acidic condition (provided by the HC1 form of the ligands) and an alcoholic solvent. Pre-association of the perrhenate and the cationic protonated ligand is most likely very important in this reduction, although the formation of a ReVIIC>3-(pmen) complex remains elusive. Curiously, while bulk synthesis of [ReOCl2(H2pmen)]Cl-CH3OH resulted in microcrystalline material containing one methanol solvate molecule, single crystals of the material contained two lattice methanol molecules. Both complexes were characterized by IR, mass spectrometry, elemental analysis, 'H NMR, and X-ray crystal structure analysis. The infrared stretching vibrations VRe=o of the cations in [ReO(bped)][Re04]H20 and [ReO(H2pmen)Cl2]Cl-CH3OH were found to be 920 and 927 cm"1, respectively. The values are typical of Re v=0 complexes and are at the lower end of the reported values.1 0'1 5 In the mass spectra of both complexes, M + parent peaks were observed. Another predominant peak present is that of the [ReO(Ci4Hi6N4)]+ fragment, indicative of the loss of two apical pendant arms of the ligands for [ReO(bped)]+, and of loss of two HC1 for [ReO(H2pmen)Cl2]+. Suitable crystals for X-ray crystallographic analysis were grown for [ReO(bped)][Re0 4 ]»CH 3 OH (from methanol) and [ReO(H 2 pmen)Cl 2 ]Cl»2CH 3 OH (from methanol/diethyl ether). The structures of the complex cations are remarkably similar 37 (Figures 3.1 - 3.4 and Table 3.1); along with [ReO(bbpen)]PF612 and a previously reported [ReO(qtp)(OCH3)2]C104,4 they represent a unique structural motif in Re v=0 chemistry.** Figure 3.1. Thermal Ellipsoid Drawing of the [ReO(H2pmen)Cl2]+ Cation in [ReO(H2pmen)Cl2]Cl-2CH3OH Showing the Crystallographic Numbering. Thermal Ellipsoids for the Non-Hydrogen Atoms are Drawn at the 50% Probability Level. ** For clarity and completeness, the structure of [ReO0}bpen)]PF6 structure is included in this discussion. 38 Figure 3.2. Thermal Ellipsoid Drawing of the [ReO(bped)]+ Cation in [ReO(bped)][Re04]-CH3OH Showing the Crystallographic Numbering. Thermal Ellipsoids for the Non-Hydrogen Atoms are Drawn at the 50% Probability Level. 39 Thermal Ellipsoid Drawing of the [ReO(bbpen)]+ Cation in [ReO(bbpen)]PF6 Showing the Crystallographic Numbering.12 Thermal Ellipsoids for the Non-Hydrogen Atoms are Drawn at the 50% Probability Level. 40 Cl(1) [ReO(H2pmen)Cl2j+ [ReO(bped)]+ [ReO(bbpen)]+ Figure 3.4. Drawing of the Coordinating Atoms in the [ReO(H2pmen)Cl2]+, [ReO(bped)]+ and [ReO(bbpen)]+ Cations. 41 Table 3.1. Selected Bond Lengths (A) and Angles (deg) in the Three Complex Cations. [ReO(H2pmen)Cl2]+ [ReO(bped)]+b [ReO(bbpen)]+d 12 Re-O(oxo) Re-N(amine 1) Re-N(amine 2) Re-X(l) Re-X(2) Re-N (pyridine 1) Re-N(pyridine 2) 1.682(3) 2.310(3) 2.286(3) 2.361(l)a 2.356(l)a 2.247(3) 2.254(3) 1.689(4)/1.687(4) 2.320(4)/2.319(4) 2.320(4)/2.320(4) 2.014(4)/2.012(3)c 2.017(5)/2.020(4)c 2.232(4)/2.232(4) 2.228(4)72.227(4) 1.689(3) 2.363(2) 2.363(2) 2.002(2)c 2.002(2)c 2.257(2) 2.257(2) X(l)-Re-O(oxo) X(2)-Re-0(oxo) X(l)-Re-X(2) 98.5(l)a 99.3(l)a 162.25(4)a 102.8(2)/102.4(2)c 101.3(2)/101.4(3)° 155.9(2)/156.1(2)c 101.73(6)c 101.73(6)c 156.5(1)° 0(oxo)-Re-N(pyl) 77.2(1) 0(oxo)-Re-N(py2) 77.7(1) N(aml)-Re-N(pyl) 67.7(1) N(am2)-Re-N(py2) 67.5(1) N(aml)-Re-N(am2) 70.0(1) 80.1(2)/79.9(2) 80.5(2)/80.7(2) 68.4(1 )/68.3(2) 68.4(2)/68.5(2) 73.2(2)773.0(2) 76.40(6) 76.40(6) 67.38(8) 67.38(8) 72.8(1) a X = CI. b two unique molecules per asymmetric unit. ° X = 0 . d symmetry operation used to generate equivalent atoms: -x+l, y, -z+'A. 42 All three metal complexes made in our group (Figure 3.4) are C 2 symmetric: the symmetry in [ReO(bbpen)]PF6 is crystallographic, and that in the other two are approximate. The rhenium center in all the complexes is seven-coordinate with a rarely observed distorted pentagonal bipyramidal geometry. The rhenium atom sits in a basal plane formed by an oxo O and four N atoms. The basal planarity indicates the degree of strain imposed by the axial ligand linkage to the ethylenediamine backbone. Specifically, the basal plane of [ReO(H2pmen)Cl2]+ is almost perfectly planar, that of [ReO(bbpen)]+ is slightly skewed, and that of [ReO(bped)]+ is significantly distorted, most likely by chelation of N C H 2 C O O " units which form five-membered rings (Figure 3.4). This planarity of the basal plane is also observed in [ReO(qtp)(OCH3)2]+, in which the axial methoxide ligands are not linked to the basal N 4 unit.4 The basal pentagon is distorted, possibly to accommodate the Re=0 linkage: the N(py) -Re-O(oxo) angles range from 76.4° to 80.7°, which are significantly larger than the N-Re-N angles (67°-73°). This is also mirrored in [ReO(qtp)(OCH3)2]C104 (in which N(py)-Re-0(oxo) ~ 77.3° and N-Re-N ~ 68.50).4 In all three structures, the two axial atoms are bent away from the oxo (O(axial)-Re-O(oxo) ~ 102° and Cl-Re-O(oxo) ~ 99°, as opposed to 90° in an ideal pentagonal bipyramid). The degree of bending seems to be unaffected by the strain imposed from the linkage of the axial atoms to the basal unit, as verified by the strain-free [ReO(qtp)(OCH3)2]C104 structure (O(axial)-Re-O(oxo) = 101.5(2)0);4 and the axial coordinating atoms do not significantly change the bending as there is < 3° difference from CI to O ligands. The oxo O, Re, and two axial atoms form an axial plane bisecting the molecule. The axial plane is perpendicular to the basal plane in [ReO(H2pmen)Cl2]+ (and 43 [ReO(qtp)(OCH3)2]+) without the ligand-imposed strain, otherwise it tilts, slightly in [ReO(bbpen)]+ and significantly in [ReO(bped)]+. The Re=0 bond lengths in the three complex cations are almost identical, verifying the remarkable similarity of the three complexes, and are within the expected range.3'15 The Re-N(pyridine) bonds are within the range of typical Re-N bonds (2.10-2.27 A in Re v O complexes),15'16 whereas the Re-N(amine) bonds exceed this range, probably because they are trans to the oxo group. The Re-O(axial) distances (-2.00 A ) of the two relevant complexes are only slightly different, and are comparable with those in [ReO(qtp)(OCH3)2]+ (~1.98 A ) . 4 This lack of variance in the axial bond lengths, along with the observation that the O-Re-O(axial) bond angle remains the same in all the complexes, suggest that the axial ligands are hardly affected by the rest of the molecule. The similarities of these complexes warrant a generalization of some interesting and indeed, surprising features of their structures: 1. The complexes all have pentagonal bipyramidal geometries with an oxo O and four neutral N atoms in the basal plane; given the spatial and electronic demands of an oxo group, one hardly expects to find it in the cluttered basal plane. 2. Without ligand imposed-strain, the basal plane is perfectly planar. 3. The anionic ligands occupy the axial positions, which are distorted away from the oxo group and the degree of bending does not seem to be affected by the nature of the ligands or the rest of the molecule. 4. The oxo O, Re and the axial ligands form an axial plane bisecting the molecule. 'H N M R spectroscopy indicates that under suitable conditions all three cations retain their solid state structures in solution. The assignments of all protons in all the complexes (Scheme 3.2) are presented in Table 3.2. All three complex cations are expected to be C 2 44 symmetric based on crystallographic results, and this symmetry was observed in their lH N M R spectra. In all three complexes 'H signals of ethylene and those of methylene adjacent to the pyridyl group split into two sets of doublets and have coupling constants consistent with geminal coupling.17 This is also true for the methylene 'H NMR resonances of the carboxylate and oxobenzene arms in the [ReO(bped)]+ and [ReO(bbpen)]+cations. Also, there is only one set of pyridine (and oxobenzene in [ReO(bbpen)]+) ! H signals. The chemical shifts of the H2 signals are more downfield than those of the other H p y signals, a feature that is typical in such systems, especially in metal complexes.13 It is clear that all the complexes retain the seven-coordinate structures in solution; this factor suggests that they are robust and that the structural motif they represent is potentially general. 45 R = H, H 2pmen bped2" bbpen 2 Scheme 3.2 Table 3.2. 'H N M R Data for the Re(V) Complexes (Scheme 3.2). Assignment [ReO(H2pmen)Cl2]Cl [ReO(bped)][Re04]-H2Ob • C H 3 O H a [ReO(bbpen)]PF6c'12 H e n 3.32 (d, 2H) 3.65 (d, 2H) 3.61 (d, 2H) H e n ' 3.47 (d, 2H) 4.18 (d, 2H) 4.26 (d, 2H) Hi 4.46 (d, 2H) 4.50 (d, 2H) 4.51 (d, 2H) Hi ' 4.88 (d, 2H) 5.22 (d, 2H) 5.36 (d, 2H) H 2 8.89 (d, 2H) 9.12 (d, 2H) 8.17 (d, 2H) H 3 7.55-7.7 (m) 7.6-7.8 (m) 7.16 (t, 2H) H 4 7.93 (td, 2H) 8.03 (td, 2H) 7.82 (td, 2H) H 5 7.55-7.7 (m) 7.6-7.8 (m) 7.61 (d, 2H) H 6 — 3.65 (d, 2H) 3.80 (d, 2H) H 6 ' — 3.77 (d, 2H) 4.63 (d, 2H) H 7 — — 6.31 (d, 2H) H 8 — — 6.95 (td, 2H) H 9 — — 6.31 (t, 2H) Hio — — 6.44 (d, 2H) " C D 3 0 D / D 2 0 / D C 1 . b C D 3 O D . c acetone-^4 46 3.4 Conclusions Oxorhenium(V) complexes with H2pmen and H2bped have been synthesized and structurally characterized. Along with [ReO(bbpen)]+ and [ReO(qtp)(OCH3)2]+, [ReO(H2pmen)Cl2] and [ReO(bped)]+ share a rare seven-coordinate structure with a distorted pentagonal bipyrarnidal geometry, which represents a novel, and potentially general structural motif in Re v=0 complexes. The solution structures of the complexes, as shown by the 'H N M R spectra, are similar to those in the solid state. 3.5 References 1) Pierreroy, J. T. M.Sc Thesis; University of British Columbia: Vancouver, 1997. 2) Connor, K. A.; Walton, R. Comprehensive Coordination Chemistry; Wilkinson, G., Gillard, R. D. and McCleverty, J. A., Ed.; Pergamon Press: Oxford, 1987; Vol. 4. 3) Rouschias, G. Chem. Rev. 1974, 74, 531. 4) Che, C. -M.; Wang, Y. P.; Yeung, K. S.; Wong, K. Y.; Peng, S. M . J. Chem. Soc., Dalton Trans. 1992,18, 2675. 5) Sugimoto, H.; Kamei, M . ; Umakoshi, K.; Sasaki, Y.; Suzuki, M . Inorg. Chem. 1996, 35, 7082. 6) Sugimoto, H.; Sasaki, Y. Chem. Lett. 1997, 541. 7) Takahira, T.; Umakoshi, K.; Sasaki, Y. Acta Crystallogr., Sect. C 1994, 50, 1870. 8) Liu, S.; Rettig, S. J.; Orvig, C. Inorg. Chem. 1991, 30, 4915. 9) Luo, H.; Rettig, S. J.; Orvig, C. Inorg. Chem. 1993, 32, 4491. 10) Luo, H.; Liu, S.; Rettig, S. J.; Orvig, C. Can J. Chem. 1995, 73, 2272. 47 11) Luo, H.; Setyawati, I.; Rettig, S. J.; Orvig, C. Inorg. Chem. 1995, 34, 2287. 12) Setyawati, I. A. Ph. D. Thesis; University of British Columbia: Vancouver, 1999. 13) Caravan, P.; Rettig, S. J.; Orvig, C. Inorg. Chem. 1997, 36, 1306. 14) Branca, M . , Checconi, P., Pispisa, B. J. Chem. Soc, Dalton Trans. 1976, 481. 15) Botha, J. M . ; Umakoshi, K.; Sasaki, Y.; Lamprecht, G. J. Inorg. Chem. 1998, 37, 1609. 16) Lock, C. J. L.; Turner, G. Acta Crystallogr. Sect. B 1978, 34, 923. 17) Silverstein, R. M . ; Bassler, G. C ; Morrill, T. C. Spectrometric Identification of Organic Compounds; John Wiley & Sons: New York, 1980. 48 Chapter 4. Chemistry of Re Complexes Containing N,N'-Bis(2-pyridylmethyl)ethylenediamine (H2pmen): Hydrolysis, Dehydrogenation, and Ternary Complexes 4.1 I n t r o d u c t i o n Numerous commercial Tc imaging agents owe their success to their chemistry in vivo. In general, the compound is distributed by perfusion but retained by biomodification.1 Thus, in a typical approach, a coordination complex is defined and its ligand is then modified to vary its metabolic biochemistry.1'2 Alternatively, ternary complexes offer the possibility of biomodification on the metal site: substitution of one ligand would most likely affect the biochemistry of the complex. To approach this idea, one needs a "primary" ligand that forms not only a stable metal complex by itself but also forms ternary complexes incorporating a "secondary" ligand. Ultimately, this design strategy can be extended to making ternary complexes with, or interacting with, various biologically interesting endogenous ligands. There is extensive literature on rhenium complexes with N-donor ligands.3"11 With appropriate halides or oxygen donor ligands, these complexes are usually cationic and stable (or inert) in air and in water in oxidation states III, IV and V. Because these N-donors are mostly neutral, their complexes commonly react with anionic oxygen donor ligands to form ternary complexes. This charge difference between N and O donors is an excellent tool in a one-pot synthesis to produce a self-assembled ternary complex. The reaction chemistry of ReL complexes (where L = N-donor ligands) is rich, and a number of these reactions have implications in nuclear medicine. Most notable reactions, for our purposes, include oxygen 49 transfer reactions,4'8'12 redox,5'13 reduction/complexation (with, for example, organohydrazines),14 and substitution with anionic ligands.5 With a suitable chelating ligand, these reactions usually do not lead to demetallation of the N-donor ligand. For instance, many air-stable Re complexes with tripodal N ligands have been made in III, IV and V oxidation states, and the aforementioned reactions are common in these complexes.5'1 1-1 3'1 5 N-donor ligands are thus good candidates for our 'primary ligands'. The chemistry of Tc/Re with 3-hydroxy-4-pyrones, a family of monoanionic bidentate O donor ligands, has been studied.16'17 These anionic O-donor ligands stabilize M v with an oxo or nitrido core, but their M1WW complexes are ill-defined, probably because of their reactivity in air and in water. It is evident that these ligands alone are unable to stabilize the metal center very well. However, this inability is an attractive feature for a 'secondary ligand' in a ternary complex: it would not displace the anchoring ligand. An added advantage of these ligands is that they are easily functionalized, providing a wealth of possible bioconjugation chemistry.18"23 H2pmen is a potentially tetradentate ligand with four N-donor atoms. Following on from a long line of amine/pyridine ligands, it promises stable Re complexes of intermediate oxidation states and presents the possibility of addition products with anionic ligands. We have previously reported a 7-coordinate complex [ReOCl2(H2pmen)]Cl (4.1, Scheme 4.1, see Chapter 3). 2 4 In this study, we investigated the hydrolysis behaviour of 4.1 and one-pot syntheses of ternary complexes from this ligand, [NH 4 ] [Re04] and 3-hydroxy-4-pyrones. 50 NH HN + 4HCI P(C 6H 5) 3 cr [NH4][Re04] H2pmen4HCl 4.1 Scheme 4.1 4.2 Experimental Materials and Instrumentation. See Chapter 2 for details. 3-hydroxyflavone was obtained from Aldrich Chemicals and was used without further purification. Maltol (2-methyl-3-hydroxy-4-pyrone), ethylmaltol (2-ethyl-3-hydroxy-4-pyrone) and kojic acid (3-hydroxy-5-hydroxymethyl-4-pyrone) were obtained from Pfizer. N,N'-bis(2-pyridylmethyl)ethylenediamine tetrahydrochloride (H2pmen»4HCl) and [ReOCl2(H2pmen)]Cl (4.1) were prepared as described in Chapter 3. 2 4 CI analysis was also performed by Mr. Peter Borda in this department. [ReOCl(Hpmen)][Re04], 4.2. [NH 4 ][Re0 4 ] (26 mg, 0.10 mmol) and H2pmen«4HCl (38 mg, 0.10 mmol) were suspended in methanol (5 mL) with heating. Triphenylphosphine (52 mg, 0.20 mmol) was added to the suspension and the mixture was heated at ~ 60°C for a few minutes, resulting in precipitation of a green crystalline material which was collected and dried under vacuum. Anal. Calcd (found) for Ci 4 Hi7ClN 4 0 5 Re 2 : C, 23.06 (23.22); H, 2.35 (2.36); N, 7.68 (7.59); CI, 4.86 (5.05). Mass spectrum (LSIMS): m/z = 551 ([ReO(Ci 4 Hi 5 N 4 )(C 3 H 9 02S)] + , where C 3 H 9 0 2 S is the matrix, thioglycerol, and C i 4 H i 5 N 4 = 51 H2pmen-3H), 443 ([ReO(Ci 4H 1 6N 4)] +). IR (cm"1, KBr disk): 940, 909 (vR e= 0 for the complex cation, and [Re04]", respectively). Single crystals suitable for X-ray crystallographic analysis were grown from a water/methanol solution of [ReOCl2(H2pmen)]Cl-CH30H (4.1) in a closed vial. [Re0 2 (H 2 pmen)][Re04], 4.3. Method A. To H 2pmen*4HCl (205 mg, 0.846 mmol) in acetonitrile (3 mL) was added sodium acetate (145 mg, 1.76 mmol) in methanol. The resulting solution was purged with nitrogen for 20 min, and [(C4H9)4N][ReOBr4] (430 mg, 0.563 mmol) in acetonitrile (5 mL) was added dropwise. The purple/black solution was stirred for 24 h at room temperature to yield a violet precipitate, which was filtered out and dried under vacuum. Method B. [NH4][Re04] (27 mg, 0.10 mmol) and H2pmen»4HCl (38 mg, 0.10 mmol) were dissolved in methanol (5 mL)/water (5 mL). Excess NaBH 4 (15 mg, 65 mmol) was added resulting in effervescence, and the mixture turned dark red. More [NH4][Re04] (31 mg, 0.11 mmol) was added and the mixture was kept at room temperature for 24 h over which time dark purple microcrystals precipitated. The precipitate was collected and dried. Method C. Crystals of [ReOCl2(H 2pmen)]Cl-CH 3OH, 4.1, were suspended in methanol in an open vial at 4°C. Green crystals of [ReOCl(Hpmen)][Re04] 4.2 formed in a few days; subsequently, dark needles of [Re02(H2pmen)][Re04] 4.3 deposited. Conversion to 4.3 was complete in a few weeks. Alternatively, 4.1 or 4.2 was treated with excess sodium acetate, and the dark red material precipitated immediately. Anal. Calcd (found) for Ci 4Hi 8N 406Re2: C, 23.66 (24.19); H, 2.55 (2.54); N, 7.88 (7.84). Mass spectrum (LSIMS): m/z = 443 ([ReO(Ci 4Hi 6N 4)] +, where d 4 H 1 6 N 4 = H2pmen-2H). IR (cm"1, KBr disk): 911, 788 (vR e= 0 52 for [Re0 4]\ and vtTans R e 0 2 , respectively). 'H NMR (DMSO-^) 8: 4.0 (d, 2H, H e n ); 4.6 (two sets of overlapping doublets, 4H, H e n and NC7/2Py); 6.1 (d, 2H, NC// 2Py); 7.8 (td, 2H, H p y ); 8.2 (td, 2H, H p y ); 8.0 (d, 2H, H p y ); 9.7 (d, 2H, H p y ) . [Re(NCS)3(Ci4Hi6N4)], 4.4 ( C 1 4 H i 6 N 4 = H2pmen-2H). [ReOCl(Hpmen)][Re04] (60 mg, 0.082 mmol) was suspended in 10 mL methanol and KNCS (15 mg, 0.15 mmol) was added. Green crystals of [ReOCl(Hpmen)][Re04] dissolved in a few minutes followed by precipitation of a white crystalline material (a mixture of K[Re0 4] and KC1). The supernatant was decanted and refrigerated, and dark purple crystals precipitated in the red solution within a day. The supernatant was decanted off and the crystals were dried. Anal. Calcd (found) for Ci 7H 1 6N 7ReS3: C, 33.99 (33.82); H, 2.68 (2.64); N, 16.32 (15.91). Mass spectrum (+LSIMS): m/z = 655 ([ReS 2(NCS)(Ci 4Hi 4N 4)(C 3H 80 2S)] +, where C 3 H 8 0 2 S is the matrix, thioglycerol), 623 ([ReS(NCS)(C 1 4Hi 4N 4)(C 3H 80 2S)] +), 565 ([ReS 2(NCS)(Ci 4H 1 4N 4)-H 20] +), 551 ([ReS 2(CS)(Ci 4Hi 4N 4)-H 20]+), 489 ([ReS 2(Ci 4Hi 4N 4)] +). IR (cm 4, KBr disk): 2081,2059 (vN=c). [ReCl(ma)(C14H,4N4)][Re04]-CH3OH, 4.5 ( C i 4 H i 4 N 4 = H2pmen-4H). [NH4][Re04] (27 mg, 0.10 mmol), H2pmen-2HC1 (38 mg, 0.10 mmol) and 2-methyl-3-hydroxy-4-pyrone (maltol, Hma, 42 mg, 0.33 mmol) were combined in methanol (3 mL) and the solution was heated for 20 h at 70°C. To the resulting hot dark green solution was added more [NH4][Re04] (27 mg, 0.10 mmol) and single crystals suitable for a structural determination precipitated upon slow cooling. The crystals were collected, washed with methanol, and dried. Anal. Calcd (found) for C 2 1 H 2 3 C l N 4 0 8 R e 2 : C, 29.08 (29.43); H, 2.67 (2.60); N, 6.46 53 (6.55); CI 4.09 (3.98). Mass spectrum (+LSIMS): m/z = 585 ([ReCl(ma)(Ci4Hi4N4)]+). IR (cm"1, KBr disk): 909 (vRe=o). [ReCl(ema)(Ci4Hi4N4)][ReC>4], 4.6 (C14H14N4 = H2pmen-4H). The synthesis was the same as that for [ReCl(ma)(Ci4Hi4N4)][Re04], except that 2-ethyl-3-hydroxy-4-pyrone (ethylmaltol, Hema, 42 mg, 0.30 mmol) was used. Anal. Calcd (found) for C2iH2iClN407Re2: C, 29.70 (29.51); H, 2.49 (2.45); N, 6.60 (6.30). Mass spectrum (+LSIMS): m/z = 599 ([ReCl(ema)(Ci4Hi4N4)]+). IR (cm"1, KBr disk): 909 (vRe=o). [ReCl(koj)(Ci4Hi4N4)][Re04]-CH3OH, 4.7 ( C i 4 H i 4 N 4 = H2pmen-4H). The synthesis was the same as that for [ReCl(ma)(Ci4Hi4N4)][Re04], except that 5-hydroxymethyl-2-hydroxy-4-pyrone (kojic acid, Hkoj, 42 mg, 0.30 mmol) was used. Anal. Calcd (found) for C2iH23ClN409Re2: C, 28.56 (28.70); H, 2.62 (2.34); N, 6.34 (6.25). Mass spectrum (+LSIMS): m/z = 601 ([ReCl(koj)(Ci4H14N4)]+). IR (cm"1, KBr disk): 909 (v R e = 0 ) . [ReCl(flv)(Ci4Hi4N4)][Re04], 4.8 ( C 1 4 H 1 4 N 4 = H2pmen-4H). [NH4][Re04] (0.027g, 0.10 mmol), H2pmen-4HC1 (0.038g, 0.10 mmol) and 3-hydroxyflavone (Hflv, 34 mg, 0.14 mmol) were combined in methanol (3 mL) and the solution was heated for 20 h at 70°C. This resulted in a dark green solution and a dark purple crystalline material. The crystals was collected and dried. Anal. Calcd (found) for C s g l f e C l r ^ C ^ : C, 36.77 (37.04); H, 2.45 (2.47); N, 5.91 (5.97). Mass spectrum (+LSIMS): m/z = 697 ([ReCl(flv)(Ci4Hi4N4)]+), 662 ([M-C1]+), 621 ([M-C 6H 4] +). IR (cm"1, KBr disk): 910 (vRe=0). 54 [ReOCl2(Ci4H1 4N4)][Re04], 4.9 (CJ4H14N4 = H2pmen-4H). [NH4][Re04] (0.027g, 0.10 mmol), H2pmen-4HC1 (0.038g, 0.10 mmol) and the HC1 salt of 2,5-diazo-N,N'-dimethyl-l,6-bis(phenylphosphinic acid) (98 mg, 0.20 mmol, see Chapter 2) were combined in methanol (3 mL) and the solution was heated for 70 h at 70°C. The resulting dark red solution was cooled to room temperature, and left in the open air for a few days over which time a yellow microcrystalline solid precipitated. The precipitate was collected. Anal. Calcd (found) for C i 4 H i 4 C l 2 N 4 0 5 R e 2 : C, 22.08 (21.68); H, 1.85 (1.93); N, 7.36 (7.19). Mass spectrum (+LSIMS): m/z = 511 ([ReOCl 2(Ci 4Hi 4N 4)] + +). IR (cm"1, KBr disk): 942, 910 (vRe=o for the complex cation, and perrhenate, respectively). 'H NMR (DMSO-£fc) 6: 4.5 (s, 4H, H e n ); 8.1 (m, 2H, Hpy); 8.8 (m, 2H, H p y ); 8.6 (d, 2H, H p y ); 9.4 (s, 2H, NC#Py); 9.6 (d, 2H, H p y ) . 55 4.3 Results and Discussion As reported in Chapter 3, when stoichiometric amounts of ammonium perrhenate, Fkpmen 4HC1 and triphenylphosphine were heated in methanol at 70°C, the solution turned immediately green, then in about an hour, it turned yellow and yielded [ReOCl2(H2pmen)]Cl (4.1).24 A green crystalline material was isolated in high yield when excess [NH4][Re04] was used and the reaction mixture was heated only for a few minutes. This green material was characterized as [ReOCl(Hpmen)][Re04] (4.2). The elemental analysis fits, the LSIMS(+) spectrum shows a prominent [ReO(pmen)]+ ([M-HC1]+) peak at m/z = 443, and Re=0 stretches of both perrhenate (vReo =909 cm"1) and the complex (vReo = 940 cm"1) are prominent in the IR spectrum. Surprisingly, this formulation suggests the formation of an amido-Re bond in a wet acidic solution (the reaction started with Fkpmen 4HC1, which rendered the solution acidic). Water stable M-amido bonds (M = Tc or Re) are rare, but the few existing examples are well known in the field of nuclear medicine. T c - 0 complexes of HM-PAO and ECD, both containing amido donors, are successful brain-imaging agents.1 Single crystals of 4.2 precipitated from a methanol/water solution of [ReOCl2(H2pmen)]Cl-CH3OH (4.1), and an X-ray crystallographic analysis of 4.2 confirmed the formulation. One oxo O, one CI, and four N atoms coordinate to the rhenium center forming a distorted octahedron, with the O and CI atoms cis (Figure 4.1 and Table 4.1). Except for a remarkable right angle formed by O-Re-Cl, the remaining coordinating atoms are squeezed towards one another and away from the oxo O atom. This distortion is consistent with other [Re=0]3+ complexes, and can be rationalized by the spatial and electronic demands of an oxo ligand, and/or strain imposed by five-membered ring chelation.1 7-2 5-2 9 An amido-Re bond is confirmed by its short length (Re-N(3) = 1.896 A) 56 and by the (almost) planar sp2 hybridized N(3). The rest of the Re-N lengths, as well as Re=0 (1.691 A) and Re-Cl lengths, are in the expected range, with Re-N(l) being longest (2.296 A), consistent with coordination trans to an oxo ligand. Cd) C(6) H(17) Cl(l) Figure 4.1. C(3) ORTEP Drawing of the [ReOCl(Hpmen)]+ Cation in 4.2, Showing the Crystallographic Numbering. Thermal Ellipsoids For the Non-Hydrogen Atoms are Drawn at the 33% Probability Level. 57 Table 4.1. Selected Bond Lengths (A) and Bond Angles (deg) in the Complex Cations in 4.2 and 4.5. [ReOCl(Hpmen)]+, 4.2 [ReCl(ma)(Ci 4H 1 4N 4)]+, 4.5 Re-O(l) 1.691(5) 2.134(4) Re-0(2) 2.034(4) Re-N(l) 2.296(6) 2.091(4) Re-N(2) 2.137(6) 2.090(5) Re-N(3) 1.896(6) 2.056(5) Re-N(4) 2.089(6) 2.102(4) Re-Cl(l) 2.464(2) 2.499(1) N(2)-C(6) 1.49(1) 1.287(7) N(3)-C(9) 1.465(9) 1.307(7) Cl(l)-Re-0(1) 90.0(2) 80.6(1) 0(2)-Re-0(l) 77.3(2) N(l)-Re-0(1) 165.6(2) 161.9(2) N(2)-Re-0(1) 98.4(3) 121.3(2) N(3)-Re-0(1) 109.6(3) 73.0(2) N(4)-Re-0(1) 102.3(2) 90.2(2) N(l)-Re-N(2) 75.1(2) 73.1(2) N(l)-Re-N(3) 82.4(3) 124.7(2) N(l)-Re-N(4) 87.3(2) 91.7(2) N(l)-Re-Cl(l) 78.2(2) 93.2(1) N(2)-Re-N(3) 81.2(3) 68.5(2) 0(2)-Re-N(l) 85.1(2) 0(2)-Re-Cl(l) 82.1(1) 0(2)-Re-N(4) 83.3(3) Re-N(3)-C(9) 120.4(5) 119.4(4) Re-N(3)-C(8) 119.8(5) 118.4(4) C(8)-N(3)-C(9) 115.0(6) 121.8(5) 58 The 'H NMR spectrum (Figure 4.2) shows that 4.2 retains its solid state structure in solution. All seventeen H atoms are seen: sixteen in the ligand backbone (a, b, c, see Scheme 4.2) and one amine H, which disappears on addition of D2O. In a COSY experiment, the amine proton is shown to couple to one of the methylene protons next to the pyridine. As expected, coupling patterns of the ethylene H atoms are complex (each couples to the other three), those ofNCH^py are simple (geminal coupling only), and those of pyridine H atoms are unique (each couples to the others on the same pyridine ring in a predictable fashion). All ' H N M R signals are well resolved, suggesting a rigid solution structure. solvent .b b c 11.0 10.0 9.0 8.0 7.0 (ppm) 6.0 4.0 4.0 3.0 2.0 Figure 4.2. H N M R Spectrum of [ReOCl(Hpmen)][Re04], 4.2, in DMSO-d6. 59 Interestingly, from a mixture of [NH4][Re04] and H2pmen-4HC1, when NaBH 4 was used as a reductant instead of triphenylphosphine, [Re02(H.2pmen)] [Re04] (4.3) was isolated as a dark crystalline solid. The IR spectrum of the complex is marked by a prominent asymmetric stretch of trans Re02 ( V R C O 2 = 788 cm"1), diagnostic of such species;30 and the LSIMS(+) spectrum shows a [M-H 20] + peak at m/z = 443. The N M R spectrum of 4.3 shows the complex to be either C2 or C s symmetric with a coupling pattern similar to that of 4.1, but the chemical shifts of the corresponding protons of 4.1 and 4.3 are markedly different, reflecting differences in their coordination environments. 4.2 Scheme 4.2 60 Conversion among 4.1, 4.2, and 4.3 is colorful and interesting (Scheme 4.2). In a methanol/water solution in air, [ReOCl2(H2pmen)]Cl (4.1, yellow) slowly converts to [ReOCl(Hpmen)][Re04] (4.2, green) and finally to [Re02(H2pmen)][Re04] (4.3, red). From seven-coordinate 4.1 to six-coordinate 4.2 involves loss of one equivalent of HC1, with formation of an amido-Re bond. An analogy can be drawn to the conjugate base catalyzed dissociative mechanism (DCb) that is accepted in the substitution of Co 1 1 1 amine complexes: a proton on an amine ligand is extracted by a base, forming an amido-Co bond with concomitant loss of a ligand (for example, a chloride), and the highly reactive 5-coordinate intermediate accepts a donor ligand, completing the substitution reaction.3 1'3 2 The proposed five-coordinate intermediate, however, has not been isolated (or observed) in Co complexes with simple amines. Complex 4.2, owing to the inertness of Re and its structurally constraining ligand, lends some support to the DCb mechanism. Following this analogy to Co chemistry and assuming that 4.2 is indeed the coordinatively unsaturated intermediate, the amido complex is expected to continue base catalyzed hydrolysis, or to revert to [ReOCl2(H2pmen)]Cl in excess HC1. The conversion from 4.2 to 4.3 can be viewed as loss of one HC1 and gain of one H 2 0 , effectively equivalent to hydrolysis. A drop of base completes the conversion from 4.1 or 4.2 to 4.3 instantly; and solutions of 4.2 or 4.3 turn yellow (to 4.1) when HC1 is added. This process is notable because it supports conjugated base catalyzed hydrolysis and provides a rare example of an intermediate thereof in a Re system.31'32 A coordinatively unsaturated complex also promises facile reactions (most likely via an initial addition step). 4.2 reacted rapidly with KSCN forming a number of products that were hard to isolate, except for one: a dark purple crystalline solid precipitated from a cold 61 reaction mixture of 4.2 and KSCN. Elemental analysis of the material fits a 1:1:3 ratio of Re:L:NCS, and indicates that the complex has no other non-hydrogen atoms. The IR spectrum shows no Re=0 stretch but strong C N stretches from thiocyanate (VC=N = 2081 and 2059 cm"1), and the frequencies of the signals do not provide any definitive evidence whether the NCS ligands are N-bonded or S-bonded.33 (+)LSIMS shows an interesting species of the formula [ReS2(NCSXCi4Hi4N4)]+-H20, in which 4 H atoms on H2pmen are missing. While it is common to observe [L-4H] peaks in the mass spectra of metal diamine complexes, they do not necessarily suggest that the formation of an imine complex should be possible on a synthetic scale. More telling evidence of imine formation comes from the 'H N M R spectrum (Figure 4.3). A total of 16 H atoms are seen, and one disappears (8 7.2-7.3, broad) on D2O exchange. Of the remaining 15, six H atom resonances spread out in the range 8 3.0-5.5 ppm (a and b, see Scheme 4.3 for assignment), a range that is expected from the eight methylene hydrogens on F^pmen: two H atoms are missing. Nine H atoms resonate between 8 7.0 and 9.3, in the range one would expect to find hydrogen atoms on unsaturated carbons, i.e. the eight pyridine H atoms. The extra hydrogen is a singlet, overlapping with half of a pyridine H doublet, at 8 8.25 ppm (c). This resonance at 8.25 ppm can only be explained by a -CH=N linkage in order to rationalize the integration of the spectrum, and the single exchangeable (amine) H in the D2O experiment. The complex is thus formulated as 4.4, a neutral seven-coordinate Re111 complex with three NCS ligands, and the N4 ligand coordinates through the two pyridine N, one intact amine N, and one imine N atoms. 62 - O '^ •NH | ^ C l [Re04]" xs KSCN 4.2 4.4 Scheme 4.3 Perhaps more convincing evidence for the formulation of 4.4, and indeed the reaction that produces it, comes from the synthesis of 4.5. 4.5 is a very dark green crystalline product that precipitated from a heated methanolic solution of [NH4][Re04], H2pmen-4HC1, and 63 maltol, and is formulated as [ReCl(ma)(Ci4Hi4N4)][Re04]-CH3OH. The formation of 4.5 is supported by the elemental analysis, agreeing with the above formulation; by LSIMS, showing a prominent NT1" peak at m/z = 585; and by IR spectroscopy, showing bands from both ligands and a characteristic uR eo (= 909 cm - 1) of [Re04]". The 'H N M R spectrum of 4.5 shows all 19 H atoms from the diimine N4 ligand and maltol, indicating that the complex is asymmetric (Figure 4.4). The H resonances of the maltolato ligand shift from that of the free pyrone, evincing its coordination (e, f, and g, see Scheme 4.4 for assignment). In addition to the four ethylenediamine hydrogen resonances (2 sets of multiplets, a) and eight pyridine H signals (well-resolved), two singlet imine resonances at 8.3 and 8.1 ppm are unmistakable (c), and are within the range of known imine complexes. c c J J L - i i U .A.,... 1. ,v,Ai A^... g A* 10.0 9.0 8.0 7.0 6.0 5.0 4.0 3.0 (ppm) Figure 4.4. J H N M R Spectrum of [ReCl(ma)(C14Hi4N4)][Re04], 4.5, in CDCI3/CD3OD. 64 Scheme 4.4 X-ray quality crystals of 4.5 precipitated from a slow cooling methanolic reaction mixture of (~NH4][Re04], H2pmen-4HC1 and maltol. Crystallographic analysis showed that the complex cation contains a seven-coordinate Re center complexed by one CI, two O and four N atoms (Figure 4.5 and Table 4.1). The coordination geometry resembles a capped trigonal prism with N(l)-N(4), N(2)-N(3), and 0(1)-C1(1) defining the vertices and 0(2) capping the N(1)N(4)C1( 1)0(1) face. The imine formation is supported by short C(2)-N(6) and C(9)-N(3) bond distances and by the planarity at the respective sp2 C and N atoms. The resulting shape of the complex cation is noteworthy: the Re center is coordinated by a flat ligand (maltolate) and by two flat ligand fragments (the conjugated imine-pyridine units of the C14H14N4 ligand). The flatness of the N-donor ligand is only disrupted at the sp carbons of the ethylenediamine unit, where it bends. Seven-coordinate Re111 complexes are common, and notable examples are the mono-capped tris-dioxime species of formula ReX(dioxime)3BR, where R is a variety of organic groups and X is a monodentate anion.34'3-65 A Tc analogues of these dioxime complexes, Tc(CDO)(CDOH)2Cl, is a successful heart imaging agent (see Scheme 1.1). Except for a long Re-Cl bond, bond distances of 4.5 are within the expected range for the respective bonds. The Re-N bond lengths vary little, from 2.056 A to 2.102 A, and are comparable to those of ReX(dioxime)3BR.3 4 The Re-0 bond lengths reflect the difference between enolato 0(2) coordination (2.034 A) and carbonyl 0(1) coordination (2.314 A). Figure 4.5. ORTEP Drawing of the [ReCl(ma)(Ci4H14N4)]+ Cation in 4.5, Showing the Crystallographic Numbering. Thermal Ellipsoids for the Non-Hydrogen Atoms are Drawn at the 33% Probability Level. 66 Two more ternary complexes with the 3-hydroxy-4-pyrones, ethylmaltol and kojic acid (4.6 and 4.7, respectively), and one with 3-hydroxy-4-flavone ( 4 . 8 ) were similarly synthesized in one-pot syntheses, indicating the general nature of this reaction (see Scheme 4.4 on p. 65). In a ! H NMR experiment where a 1:1:1 mixture of [NH4][Re04], H2pmen-4HC1, and maltol was heated, only the ternary product was formed, and all of the ligands were consumed. Considering the complexity of the system, this result is quite remarkable. Although the mechanism is unclear, the reaction can be characterized as a dehydrogenation, in which two amines on I-^ pmen are oxidized to imines with a concomitant four electron reduction of perrhenate (Re v n) to Re111. Dehydrogenation of amines in metal complexes is common, especially in macrocyclic amines, but the metals usually retain their oxidation state. 3 6 - 3 8 Amine dehydrogenation by a metal in a high oxidation state with concomitant reduction of the metal has been reported, but is much less common.3 8 An analogy can also be drawn to 2-hydrazinopyridine (HYPY) reactions with [M04]" (M=Tc/Re), where two H Y P Y molecules reduce [MO4]" to M m with the hydrazines being dehydrogenated and complexed to the metals in their oxidized forms, essentially, equivalent to dehydrogenations.39 We were curious about the role of the 3-hydroxy-4-pyrones in this reaction and the possibility of making ternary complexes with other anionic ligands. It was found that the 3-hydroxy-4-pyrones are crucial in this dehydrogenation, and that they most likely facilitate the r reaction by thermodynamically stabilizing, or trapping, the ternary complex. When [NH4][Re04] reacted with H2pmen-4HC1 in the absence of a 3-hydroxy-4-pyrone, a small amount of very dark red, intractable material was produced over a few days, with most of the 67 starting amine left unreacted. Other co-ligands, such as cysteine, glycine, iminodiacetic acid, and catechol, were tried instead of a 3-hydroxy-4-pyrone, and the reactions produced this mysterious dark red material with variable success. The formation of the red complex (or complexes) was somewhat clarified when H2ppme-2HC1, an N2O2 ligand, was used in the reaction. (Scheme 4.5) The dark red material was produced in sufficient quantity so that when it was left in the open air in a methanol suspension, a yellow crystalline solid was produced, and was characterized as [ReOCl2(Ci4Hi4N4)][Re04] (4.9). The production of this complex can be rationalized by the following. The dehydrogenation does occur in all of the above reactions, and it produces the dark red product(s); the red material presumably contains the Rem(Ci4Hi4N4) unit, and oxidizes to 4.9 in air. 4.9 was formulated on the basis of the elemental analysis, the IR spectrum (vRe=o = 942, 910 cm"1 for the complex cation and perrhenate, respectively), the mass spectrum (+LSIMS, with a prominent [M] + peak), and the ! H N M R spectrum (C2V symmetric with en H atoms in dynamic exchange). In light of the 7-coordinate structural motif in Re v=0 complexes proposed in Chapter 3, the structure of 4.9 can be predicted.24 2+ C 6 H 5 4.9 H 2 p p m e - 2 H C 1 Scheme 4.5 Facile reduction of perrhenate under mild conditions is useful in radiopharmaceutical applications. Production of 4.1-4.3 and 4.5-4.8 should translate readily to pharmacological 68 "kit" conditions (see Chapter 1 for detailed explanation).40 The hydrolysis behavior of 4.1-4.3 hints that a few exchanging complexes may be present in a solution, but it also promises interesting biochemistry. Reactivity of 4.5-4.8 is yet to be explored, but preliminary results indicate that the compounds have curious reactions with endogenous ligands. 4.4 Conclusions A number of Re complexes with fkpmen have been made. Complexes 4.1, 4.2 and 4.3 are related by hydrolysis/HCl substitution. Complex 4.2 was structurally characterized and found to contain a water-stable amido-Re bond. Dehydrogenation of the N-donor ligand from amine to imine with concomitant two-electron reduction of the Re center occurs readily in these systems. With suitable co-ligands such as 3-hydroxy-4-pyrones, ternary complexes 4.5-4.8 can be made from |TS"H4][Re04], H2pmen-4HC1 and pyrones in one-pot syntheses. 4.5 was structurally characterized, and the syntheses of these complexes may prove useful in nuclear medicine. Now that we have established this series of complexes and their ready formation under 'kit-friendly' conditions, we are curious about the interaction of the ternary complexes with endogeneous ligands and the biochemistry of the complexes, i.e. testing their built-in reaction sites in living systems. 69 4.5 References 1) Jurisson, S.; Desilets, C. P.; Jia, W.; Ma, D. Chem. Rev. 1993, 93,1137. 2) Jurisson, S. S.; Lydon, J. D. Chem. Rev. 1999, 99, 2205. 3) Rouschias, G. Chem. Rev. 1974, 74, 531. 4) Brown, S. N.; Mayer, J. M . Inorg. Chem. 1992, 31,4091. 5) Bohm, G.; Wieghardt, K.; Nuber, B.; Weiss, J. Inorg. Chem. 1991, 30, 3464. 6) Paulo, A.; Domingos, A.; Santos, I. Inorg. Chem. 1996, 35, 1798. 7) Paulo, A.; Domingos, A.; Marcalo, J.; de Matos, A. P.; Santos, I. Inorg. Chem. 1995, 34, 2113. 8) Brown, S. N.; Mayer, J. M . J. Am, Chem. Soc. 1996,118, 12119. 9) DuMez, D. D.; Mayer, J. M . Inorg. Chem. 1998, 37, 445. 10) Sugimoto, H.; Kamei, M . ; Umakoshi, K.; Sasaki, Y.; Suzuki, M . Inorg. Chem. 1996, 35, 7082. 11) Sugimoto, H.; Sasaki, Y. Chem. Lett. 1997, 541. 12) Abu-Omar, M . M . ; Khan, S. I. Inorg. Chem. 1998, 4979. 13) Gisdakis, P.; Antoczak, S.; Rosch, N. Organometallics 1999,18, 5044. 14) Rose, D. J.; Maresca, K. P.; Nicholson, T.; Davison, A.; Jones, A. G.; Babich, J.; Fischman, A.; Graham, W.; DeBord, R. D.; Zubieta, J. Inorg. Chem. 1998, 37, 2701. 15) Takahira, T.; Umakoshi, K.; Sasaki, Y. Acta Crystallogr., Sect. C 1994, 50, 1870. 16) Archer, C. M . ; Dilworth, J. R.; Jobanputra, P.; Harman, M . E. ; Hursthouse, M . B.; Karaulov, A. Polyhedron 1991,10, 1539. 17) Luo, H.; Rettig, S. J.; Orvig, C. Inorg. Chem. 1993,32,4491. 70 18) Finnegan, M . M . ; Rettig, S. J.; Orvig, C. J. Am. Chem. Soc. 1986,108, 5033. 19) Nelson, W. O.; Rettig, S. J.; Orvig, C. J. Am. Chem. Soc. 1987,109, 4121. 20) Finnegan, M . M ; Lutz, T. G.; Nelson, W. O.; Smith, A.; Orvig, C. Inorg. Chem. 1987, 2(5,2171. 21) Matsuba, C. A.; Nelson, W. O.; Rettig, S. J.; Orvig, C. Inorg. Chem. 1988, 27, 1045. 22) Nelson, W. O.; Rettig, S. J.; Orvig, C. Inorg. Chem. 1989, 28, 3153. 23) Zhang, Z.; Rettig, S. J.; Orvig, C. Inorg. Chem. 1991, 30, 509. 24) Xu, L.; Setyawati, I. A.; Pink, M. ; Young, V. G. J.; Patrick, B. O.; Rettig, S. J.; Orvig, C. Inorg. Chem. in press. 25) Herrmann, W. A.; Rauch, M . U.; Artus, R. J. Inorg. Chem. 1996, 35, 1988. 26) Gilli, G.; Sacerdoti, M . ; Bertolasi, V.; Rossi, R. Acta Cryst. 1982, B38, 100. 27) Banbery, H. J.; McQuillan, F.; Hamor, T. A.; Jones, C. J.; McCleverty, J. A. J. Chem. Soc. Dalton. Trans. 1989, 1405. 28) van Bommel, K. J. C.; Verboom, W.; Kooijman, H.; Spek, A. L.; Reinhoudt, D. N . Inorg. Chem. 1998,37,4197. 29) Luo, H.; Liu, S.; Rettig, S. J.; Orvig, C. CanJ. Chem. 1995, 73, 2272. 30) Fergusson, J. E. Coord. Chem. Rev. 1966, 459. 31) Tobe, M . The Role of Ion-Association in the Substitution reactions of Octahedral Complexes in Non-Aqueous Solution; Gould, R. F., Ed.; American Chemical Society: Washington, D. C , 1965, pp 24. 32) Basolo, F.; Pearson, R. G. Mechanism of Inorganic Reactions; 2nd ed.; Wiley: New York, 1967. 71 33) Nakamoto, K. Infrared and Raman Spectra of Inorganic Coordination Compounds; John Wiley & Sons: New York, 1986. 34) Jurisson, S.; Halihan, M . M. ; Lydon, J. D.; Barnes, C. L.; Nowotnik, D. P.; Nunn, A. D. Inorg. Chem. 1998, 37, 1922, and references therein. 35) Al-Shihri, A. S. M ; Dilworth, J. R.; Howe, S. D.; Silver, J.; Thompson, R. T. Polyhedron 1993,12, 2297. 36) Mahoney, D. F.; Berttie, J. K. Inorg. Chem. 1973,12, 2561. 37) Brown, G. M . ; Weaver, T. R.; Keene, F. R.; Meyer, T. J. Inorg. Chem. 1976,15, 190. 38) Bernhard, P.; Bull, D. J.; Burgi, H. B.; Osvath, P.; Raselli, A.; Sargeson, A. M . Inorg. Chem. 1997, 36, 2804. 39) Hirsch-Kuchma, M . ; Nicholson, T.; Davison, A.; Davis, W. M . ; Jones, A. G. Inorg. Chem. 1997, 36, 3237. 40) Technetium, Rhenium and Other Metals in Chemistry and Nuclear Medicine; S. G. Editorali: Padova, 1999; Vol. 5. 72 Chapter 5. Lanthanide Chemistry with Bis{[bis(carboxymethyI)amino]methyl}phosphinate: What Does an Extra Phosphinate Group Do to EDTA? 5.1 I n t r o d u c t i o n Lanthanides are important in the field of medicinal chemistry.1'2 The most successful examples are Gd complexes, which are widely used commercially as contrasting agents in magnetic resonance imaging (MRI).3'4 A 1 5 3 Sm complex (Sm-EDTMP, see Scheme 5.1) is commercially available for palliation of bone pain, and complexes of a number of metalloradionuclides are under clinical and pre-clinical investigation as radiotherapeutic agents.5'6 Because both the free ligands and the metal ions are toxic, and the complexes are not, these drugs have to be thermodynamically stable and/or kinetically inert. Ligand design for these complexes, therefore, typically employs a polyamine backbone incorporating anchoring groups such as carboxylate, phosphonate and (to a lesser extent) amides. The benchmark for these ligands is DTP A; E D T A complexes are often cited for comparison but are not stable enough under the challenging conditions found in vivo (Scheme 5.1).4 In a continuing quest for new chelating ligands,7 - 1 3 we rediscovered an old ligand whose sheer resemblance to E D T A demands attention. As part of a series of amine phosphinate compounds, X T was first synthesized by Maier, 1 4 and was reported again while our study was under way.1 5 Amine phosphinates are generally less coordinating at neutral pH than amine carboxylates or amine phosphonates because of the low basicity of the phosphinates.16'17 They are, however, derivatives of secondary phosphines and as such they can act as linkers, rather than only as terminal ligating groups, a limitation of carboxylates 73 and phosphonates. Interestingly, X T incorporates an easily prepared R2NCH2P(0)(OH)CH2NR2 motif. While amines are poor donors for lanthanides by themselves, with carboxylate anchors they contribute to the overall stability of the complexes.18 By the same argument, we are interested in how a phosphinate flanked between amines would contribute to the binding of a well-studied amine carboxylate system. In this chapter, the chemistry of X T with the lanthanides is reported. Two questions are essential in this study: whether or how the phosphinate group coordinates and how it contributes to the overall stability of the system. Co chemistry was studied for comparison. Some interesting protonation behaviour of the lanthanide complexes was also noted. HOOC < > COOH COOH HOOC H 2 0 3 P \ 7 P 0 3 H 2 P 0 3 H 2 H 2 Q 3 P H 4EDTA / • pN ~ ' N r \ HOOC < 7 C O O H COOH HOOC HgEDTMP H 4 T M D T A HOOC ^ C 0 ° H C O O H 0 0 0 " 7 > COOH / . _ . . C O o HOOC <( 0 H > C O O H COOH HOOC H 5 D T P A H 5 XT Scheme 5.1 7 4 5.2 Experimental Materials and Instrumentation. See Chapter 2 for details. Lanthanide starting materials, lanthanide atomic absorption standards, iminodiacetic acid, phosphinic acid (50%), NaOD (40%), DC1 (12 M) and formaldehyde (37%), were obtained from Aldrich, Alfa or Fisher Chemicals arid used without further purification. NaOD and DC1 solutions were used to adjust pD of the N M R solution. pD values were calculated by adding 0.4 to the pH meter reading.19 H 5XT'HCl'0.5H 2O. This synthesis is slightly modified from literature preparations.14'15 Iminodiacetic acid (2.7 g, 20 mmol) and a 50% solution of H3PO2 (1.3 g, 10 mmol) were combined in 6 M HC1 (4.0 mL) and heated to reflux. An aqueous solution (37%) of formaldehyde (3.2 g, 40 mmol) was added drop-wise to the mixture and the reflux continued overnight. A white precipitate formed when the reaction mixture was cooled to room temperature. This precipitate was collected by filtration, washed with methanol, and dried under vacuum. The filtrate was concentrated and a second crop of precipitate was recovered by filtration. This material was subsequently redissolved in a minimum amount of 6 M HC1, the product was precipitated with methanol, and similarly collected. Anal. Calcd (found) for Ci 0Hi9ClN 2O 1 0 .5P: C, 29.90 (29.91); H, 4.77 (4.96); N, 6.97 (6.73). 3 1 P N M R (D 20), 8: 17. ' H N M R (D 20), 8: 4.2 (8H, s, NCH 2 COO), 3.6 (4H, d, NCH 2 P). K3[Co(XT)]-2H20. H 5XT-HC10.5H 2O (0.157 g, 0.392 mmol) and CoCl 2 -6H 2 0 (0.110 g, 0.401 mmol) were dissolved in water (10 mL)/methanol (10 mL). A K O H solution (1 M) 75 was added drop-wise in order to neutralizethe solution (pH paper). More methanol was then added to the solution to precipitate the pink product, which was filtered out and dried under vacuum to yield 0.20 g (91%). Single crystals were grown by diffusing methanol (~1 mL) into an aqueous solution (~1 mL) of the product (10 mg). Anal. Calcd (found) for C i 0 H i 6 C o K 3 N 2 O i 2 P : C, 21.32 (21.24); H, 2.86 (3.03); N, 4.97 (4.79). Mass spectrum (-LSIMS): m/z = 450 ([KCo(XTH)]"), 488 ([K2Co(XT)]-). IR (cm"1): v C Oasym = 1613 cm"1, vcosym = 1395 and 1383 cm"1, vpoasym = H64 cm"1 vPosym = 1042 cm"1. K 2 [Co(XT)] for N M R study. K 3[Co(XT)]-2H 20 (50 mg) was dissolved in H 2 0 (10 mL) and charcoal was suspended in the solution. A few drops of H2O2 were added and the suspension was stirred overnight. The solution was filtered and concentrated; the purple residue was dissolved in water and precipitated with methanol. The precipitate was collected and dried under vacuum. 3 1 P {•H} NMR (D 20) 8: 12. lH NMR (D 20) 8: 4.5 (2H, d, 19 Hz), 4.0 (2H, d, 17 Hz), 3.7 (2H, d, 19 Hz), 3.6 (2H, d, 16 Hz), 3.0 (4H, 2 sets of overlapping doublets). General synthesis of M 2 [Ln(XT)] (H 2 0) x ( M = K , Na). H 5 XT-HC10 .5H 2 O (40 mg, 0.10 mmol) and LnCl 3 -xH 2 0 (Ln = La, Ho) or Ln(N0 3) 3-xH 20 (Ln - La, Sm, Gd, Lu) (0.10 mmol) were dissolved in 2 mL water. NaOH (1 M , in the case of LaCl 3-xH 20) or 1 M K O H (for the rest of the experiments) solution was added drop-wise until the solution pH was 6 - 8 (pH paper). The solution was then concentrated. To the residue was added a few drops of water; then methanol (3 - 5 mL) and acetone (5-10 mL) were added in order to precipitate the product, which was isolated by centrifugation. The residue was dissolved in methanol, precipitated with acetone and centrifuged again. The product was dried under vacuum. The 76 IR spectra of all the lanthanide complexes were mostly identical, some with additional V N O bands (Figure 5.1). Figure 5.1. Representative IR spectrum (K 2[Gd(XT)](H20)4) of Ln-XT and Ln-HXT Complexes. L , , , _ , _ , 1 , — — r-" 4000 3500 3000 2500 2000 1500 1000 500 wavenumber (cm-1) Na 2[La(XT)](H 20) 3. Anal. Calcd (found) for CioHi 8LaNa 2N 2Oi3P: C, 20.35 (20.61); H, 3.07 (3.34); N , 4.75 (4.60). 31P{'H} (D 20) NMR 5: 44, 39. 77 K2.5[Sm(XT)](H20)5(N03)o.5- Anal. Calcd (found) for C10H22K2.5N2.5Oi6.5PSm: C, 16.67 (16.52); H, 3.08 (3.18); N, 4.86 (4.69); IR (additional peak): v N o 3 free = 1385 cm - 1. 31P{'H} (D 20) N M R 8: 57 (from pD 2.3 -10.0). K 2[Gd(XT)](H 20) 4. Anal. Calcd (found) for CioH2oGdK2N2Oi4P: C, 18.23 (18.56); H, 3.06 (3.51); N, 4.25 (4.13). Ca[Ho(XT)](H20)7. The synthesis is the same as for the other Ln complexes except that excess CaCb was added into the reaction mixture. Anal. Calcd (found) for CioH26CaHoN2Oi7P: C, 17.60 (17.37); H, 3.84 (3.70); N, 4.11 (3.93). K 3[Lu(XT)](N0 3XH20)3. Anal. Calcd (found) for CioHi 8K 3LuN 3Oi6P: C, 15.81 (15.75); H, 2.39 (2.21); N, 5.53 (5.36); IR (additional peak): v N o 3 free = 1385 cm*1. 3 1P{ !H} (D 20) N M R 8:44 (from pD 2.3-10.0). G e n e r a l synthesis o f K [ L n ( H X T ) ] x H 2 0 . H 5XT-HC10.5H 2O (40 mg, 0.10 mmol) and Ln (N03 ) 3 -xH 2 0 (Ln = Nd, Gd, Lu) or DyCl 3 -xH 2 0 (0.10 mmol in any case) were dissolved in 2 mL water. K O H solution (1 M) was added drop-wise until the solution pH was 6-8 (pH paper). Concentrated hydrochloric acid was then added to the neutral aqueous solution until pH = 2-3. Methanol (3-5 mL) was then added in order to precipitate the product, which was isolated by centrifugation. The residue was washed with methanol and dried under vacuum. K[La(HXT)](H 20) 2.5. Anal. Calcd (found) for CioHi 8KLaN 2Oi 2 .5P: C, 20.88 (21.06); H, 3.15 (3.27); N , 4.87 (4.54). K[Nd(HXT)](H 20) 2.5. Anal. Calcd (found) for CioH 1 8 KN 2 NdOi 2 . 5 P: C, 20.69 (20.86); H, 3.12(3.47); N , 4.83 (4.66). 78 K[Gd(HXT)](H 20) 5. Anal. Calcd (found) for Ci 0H23GdKN 2Oi 5P: C, 18.81 (18.66); H, 3.63 (3.60); N, 4.39 (4.28). K[Dy(HXT)](H 20) 5. Anal. Calcd (found) for Ci 0H 23DyKN 2Oi 6P: C, 18.65 (18.76); H, 3.60 (3.43); N, 4.35 (4.08). Na[Lu(HXT)](H20)3(CH3COOH). To a methanolic solution (5 mL) of Lu(OOCCH 3)3(H 2 0)4 (42 mg, 10 mmol) and H 5XT-HC1 (40 mg, 0.10 mmol) was added sodium acetate (50 mg, 0.61 mmol). Acetone was added liberally to this solution in order to precipitate the product. The precipitate was isolated by centrifugation. The solid residue was redissolved in methanol, precipitated out with acetone addition, and the mixture centrifuged again to deposit the product which was dried under vacuum. Anal. Calcd (found) for Ci 2 H 2 3LuN 2 NaOi 5 P: C, 21.70 (22.01); H, 3.49 (3.55); N, 4.22 (4.23). 3 1 H (D 20) N M R 5: in addition of the overlapped multiplet from the complex, 2.2 (s, C//3COOH). Potentiometric Equilibrium Measurements. Experiments were performed with a Fisher Acumet 950 digital pH meter or an Orion E A 920 digital pH meter, and a Metrohm Dosimat automatic buret (Dosimat 665), both interfaced with a PC. A Metrohm 6.0222.100 (combination) electrode was used. All pH readings were calibrated daily by a titration with HC1. A Gran plot 2 0 was made using this data set, and the working slope and intercept of a plot of mV (measured) vs. pH (calculated) were determined and applied to the measurements of that day (pH = -log[H+] herein). The NaOH concentration (0.1 - 0.2 M) was determined by titrating against potassium hydrogen phthalate. The temperature of the solutions was kept constant at 25.0 ± 0.1 °C by a Julabo circulating water bath. The ionic strength was fixed at 0.16 M NaCl. Each batch of ligand stock solution was titrated to determine its concentration. 79 Lanthanide solutions were prepared by appropriate dilutions of atomic absorption standard solutions (Aldrich). The final solution under study typically contained 0.5 - 2 mM of each of the ligand and the metal, with a slight excess (0.1 - 0.3 times) of the ligand. Three to ten titrations were performed for each metal-ligand system and for the free ligand; data were collected in the range 2 < pH < 11. Titrations with 2:1 and 0.5:1 ligand to lanthanum ratios were performed but only 1:1 species were found at the concentrations under study. Equilibration times between 1 and 7.7 minutes were used to ensure that equilibrium was achieved. The titration data were processed with SUPERQUAD. 2 1 At least 50% of the metal was not bound to the ligand at the starting pH. Only M L and M H L species were found in the range 2 < pH <11, and inclusion of other species worsened the fit. The errors reported represent the discrepancies amongst the same metal titrations, which are significantly larger than the error of the fitting. 80 5.3 Results and Discussion H 5 X T H C I reacts with C0CI2H2O in the presence of K O H forming K3[Co(XT)]-2H20, which was characterized by EA, IR spectroscopy, LSIMS, and X-ray crystallography. Single crystals of K3 [Co(XT)] -2H 20 were grown by diffusing methanol into an aqueous solution of the complex. The complex anion is C 2 symmetric, with the ligand coordinating to the Co center via two N and four carboxylate O atoms (Figure 5.2). The phosphinate is not coordinated. The coordination geometry is distorted octahedral with N(1)N(1)*0(3)0(3)* forming a rectangular equatorial plane and 0(1) and 0(1)* occupying the axial positions. The bond lengths are comparable to those in similar Co 1 1 species (Table 5.1) 2 2> 2 3 Table 5.1. Selected Bond Lengths (A) and Angles (deg) in [Co"(XT)]3". Co-O(l) 2.051(15) Co-0(3) 2.081(2) Co-N(l) 2.165(2) N(l)-Co-0(3) 80.46(2) 0(3)-Co-0(3)* 98.09(9) N(l)-Co-N(l)* 101.24(9) 0(l)-Co-0(l)* 169.19(10) 0(l)-Co-0(3) 92.45(6) 0(l)-Co-0(3)* 94.63(6) 0(l)-Co-N(l)* 91.42(7) * Symmetry operation -JC, l-y, z. 81 Figure 5.2. ORTEP Drawing of the [Co"(XT)]3" Anion in K 3[Co(XT)]-2H 20. The molecule lies on a crystallographic C 2 axis through Co(l) and P(l). Thermal Ellipsoids for the Non-Hydrogen Atoms are Drawn at the 33% Probability Level. 82 X T forms complexes with a number of lanthanides. They are synthesized by mixing the respective metal salt and the ligand, adjusting the pH with a base, and precipitating the complex with organic solvents. The series has been characterized by EA, IR spectroscopy (Figure 5.1), and in the case of La, Sm and Lu, P NMR spectroscopy. Solid Ln-XT complexes have been isolated in both deprotonated and monoprotonated forms, consistent with the findings in solution (vide infra). The formation of the monoprotonated samples is unmistakable; they either showed drastically different solvation behaviour from the respective deprotonated species (La, Nd and Gd complexes of X T are hygroscopic when deprotonated but sparingly soluble in water when protonated), or when they dissolved in water they buffered the solution at low pH (Dy and Lu complexes). Protonated E D T A complexes with divalent and trivalent metals have been studied extensively and shown to exist both in solution and in solid state.23"25 In most reported examples one carboxylate was protonated and a non-coordinating COOH group was evident in the IR spectra. The IR spectra of all the Ln-XT complexes, protonated or not, are virtually identical, indicating a close structural relationship across the series (Figure 5.1). Lacking an X-ray structure, one can only speculate that one of the carboxylates coordinates in a protonated form, and the proton is stabilized by extensive hydrogen bonding. Similar [Ln(HEDTA)] complexes have been reported for late lanthanides, and protonated carboxylate coordination was postulated from IR evidence.26 A vanadium complex (amavadin) containing a bond between a neutral carboxylic acid to V has been structurally characterized 2 7 The numbers of included solvents or salts in these solid materials (from EA) seems to be random. The difference in these solid state complexes lies in the additional coordination 83 sites on the metal and how they interact with each other, the solvent and small ligands, i.e. the packing of the solid material. While structurally characterized Ln-EDTA complexes have been shown to be di- or tri-hydrates,28 mono-hydrated and non-hydrated species form from extensive drying.2 6 Thus drying affects packing of the solid material and, consequently, coordination behaviour of the complex changes. With a potentially hexadentate ligand (so that the metal coordination sphere is unsaturated) it has been shown crystallographically that the formation of mono-, di-, and tri- nuclear species is very sensitive to the size of the lanthanide atom.2 9 It was rationalized on the basis of the difference of the coordination number amongst the lanthanides and the resultant packing of the crystals. Following this argument, one expects that Ln-XT complexes are possibly oligomeric in the solid state. The lack of detectable ionized species in the LSIMS experiments is consistent with this speculation. Also, in the 3 1 P NMR experiments of the La-XT and Lu-XT complexes, extra peaks appear when the solutions are concentrated. (Concentration dependence is a good indication that the species is more than mononuclear.) When K 3[Co(XT)]-2H 20 is oxidized by H 2 0 2 , a Co i n -XT complex is obtained. The 3 1 P and 'H NMR experiments showed that the ligand coordinates to the Co 1 1 1 in a similar II 31 fashion to that found in the Co analogue. One resonance at 12 ppm was seen in the P N M R spectrum. Coordination normally causes a significant downfield shift in phosphinate 3 1 P N M R signals: this 12 ppm signal indicates that, most likely, the phosphinate does not coordinate to the metal. The 'H NMR spectrum showed six sets of doublets, and coupling constants are consistent with geminal coupling, implying that the complex has reduced symmetry, most likely the C 2 symmetry of the [Co nXT] 3" is retained in its Co 1 1 1 analogue . 84 •2 1 The P NMR spectra of the Lu and Sm complexes showed single resonances at 44 and 57 ppm, respectively, indicating a single species in each solution. The drastic difference in 3 1 P resonances between the Co 1 1 1 and the lanthanide complexes infers that the phosphinate in Ln-XT is coordinated and that in [Co(XT)]2" is not. Chemical shift difference between Lu-X T and Sm-XT is attributed to differences in magnetic properties of the respective metals -Sm being paramagnetic. The 'H NMR spectrum showed sharp but overlapped signals with clear geminal coupling, corresponding to a rigid solution structure. The ' H N M R resonances are more numerous and more spread out in the Sm-XT complex, and they are not as simple as those in C o m - X T , evincing that an asymmetric diastereomer exists in each solution. Two species are observed in the 3 1 P NMR spectrum of the La-XT complex, regardless of pD. Corresponding to the solid-state result, the protonated species were observed at low pD. The single 3 1 P N M R signal remains unchanged but the ! H NMR signals broadened at pD 2.4, indicating that the aminophosphinate portion of the molecule is hardly affected by the protonation but that the carboxylates are, consistent with the expected protonation of the latter. The step-wise protonation constants (pKn, equation [5.1]) of X T are remarkably similar to those of EDTA, less so to those of T M D T A (Table 5.2). H n - i X T + H = H nXT** K n [5.1] The ionic strength of the solutions was fixed with 0.16 M NaCl in this study, but E D T A and a number of polyaminocarboxylate ligands are known to bind sodium weakly at basic pH, Charges of the species are omitted for simplicity in the equations and pertinent discussion. 85 thus the protonation constants reported herein should be viewed as apparent, or conditional, constants specifically pertinent to isotonic conditions. Protonation constants of this ligand and its metal binding stability constants (with divalent metal cations) were published when our study was under way.3 0 Their experiments were performed in tetramethyl ammonium chloride and their first proton association constant was about 10 times higher than ours, a difference that was also noted in the E D T A system.31 Table 5.2. Step-Wise Protonation Constants (25°C). X T a X T b E D T A C E D T A d T M D T A d log Ki 9.27(5) 10.27 9.52 10.37 10.27 logK 2 6.56(6) 6.76 6.14 6.19 7.91 logK 3 2.84(13) 2.86 2.88 2.87 2.67 logIC, 2.32(18) 2.26 2.11 2.26 2.00 a u. = 0.16 M NaCl, this study, b u, = 0.1 M [(CH 3) 4N]C1, 3 0 c ji = 0.12 M N a C l , 3 2 V = 0.1 M[(CH 3 ) 4 N]C1. 3 3 The complex formation constants of a few lanthanides and X T ( K M L ) were determined potentiometrically. The following equilibria (equations [5.2] and [5.3]) were included to describe the system. M + H L M H L K M H L [5-2] M + L ==— M L K M L [5.3] 86 Inherent in the above two equations, is the K a of M H L (equation [5.4]) and the conventional P M H L (equation [5.5]). M L + H M H L K a [5.4] M + H + L — - M H L P M H L [5.5] Consistent with solid state and NMR evidence, M H L complexes were found at low pH, and were completely deprotonated to M L species by pH 4.5 (Figure 5.3). Figure 5.3. Speciation Diagram of the Gd-XT system ([Gd] = [XT] - 1 mM, T = 25°C, u. = 0.16 M NaCl). 100 - i _____—— 2 3 4 5 6 PH Interestingly, log K a values of these M H L species decrease across the series with the exception of Ho. (The latter data set has larger error which invalidates the comparison.) The 87 metal complex becomes a better acid going across the series from La to Yb. Along with the solid state IR evidence that there is no free carboxylic acid donor group in the M H L species, this result points to a curious behaviour of these protonated complexes. Instead of dissociation of one carboxylate upon protonation, the carboxylate coordinations remain intact, and the proton is distributed around the complex, probably close to the carboxylates. In other words, the complex, rather than a single group, acts as an acid. Table 5.3. Logarithms of the Formation Constants for Complexation of Ln 1 1 1 with X T (T = 25°C, u. = 0.16 M NaCl). log K M H L log K M L logK a log P M H L La 7.9(3) 13.0(3) 4.14(6) 17.14(22) Nd 8.64(15) 14.64(11) 3.27(2) 17.91(12) Sm 9.30(11) 15.60(15) 2.97(5) 18.57(8) Gd 9.22(16) 15.73(25) 2.76(6) 18.48(15) Ho 8.8(4) 15.8(5) 2.2(5) 18.1(3) Yb 8.73(20) 15.56(14) 2.44(12) 18.01(18) It is difficult to compare the formation constants of Ln-XT complexes with those of other systems because of the aforementioned Na-XT association. The stability constants determined will be slightly lower than the actual values. However, pM, (equation [5.6]) an often quoted conditional quantity, is less biased in this regard.34 pM = -log [Mfree] [5-6] 88 The free metal (aqua or hydroxo species) concentration depends on the pH, as well as on metal ion and ligand concentrations, thus pM nullifies the inherent dependence of K on the basicity of the ligands in potentiometric determinations, and is often used to compare the thermodynamic stability of different chelates at a fixed pH. The expression of basicity correction intrinsic in pM is Keff (equation [5.7]), which is often used on its own in simple systems.4 Ke f f= K/(l + Ki[H +] + K 1 K 2 [ H + ] 2 +...+ K 1 K 2 . . . K n [ H + ] n ) [5.7] Sodium binds to X T at high pH, hence only the two basic pK a values differ depending on whether the measurements were done with or without Na + . The lanthanides form complexes under acidic conditions in a potentiometric determination: the M L complexes are completely formed by pH 4. It is reasonable to assume that the two events (formation of La and Na complexes) are independent of each other. Thus one can view that the two systems are different mostly because of the pK a's of the ligand, which is corrected for by Keff or pM. For example, log K A I - E D T A values are 15.3 and 16.5, determined in 0.12 M N a C l 3 2 and in 0.10 M [(CH 3) 4N]C1, 3 3 respectively, and their pM's are 14.2 and 14.5. To completely reconcile studies done in different media, one has to take deviations caused by sodium binding into account (log K N A - E D T A = 1 -86, equation [5.8]).33 Keff = K/(l + K,[H +] + KiK 2 [Pf ] 2 +...+ K 1 K 2 . . . K n [ H + ] n + K N a L[Na +]) [5.8] 89 As long as its contribution remains very small compared with the basicity correction, as is the case when the set pH value is significantly smaller than the first pK a of the ligand, pM values can be compared. The advantage of pM is that it reconciles most of the existing studies done in different supporting electrolytes, ([(CH3)4N]C1, KC1 and NaCl at similar ionic strengths), but because this comparison relies heavily on the aforementioned assumptions, it has to be treated cautiously. Comparison of the pM values of the Gd-XT system with those of the Gd-EDTA and Gd-TMDTA systems shows that the Gd-XT complex is on a par with the Gd-EDTA complex but a few orders of magnitude more stable than the Gd-TMDTA complex. Thus the addition of a phosphinate group has a similar effect on the stability of the Gd complexes as going from a six-membered chelate ring to a five-membered ring (Table 5.4 and Figure 5.4). Table 5.4. pM Values of Relevant Systems ([L] = 1(T5 M , [M] = 10"6 M , pH = 7.4, T = 25°C). E D T A 3 T M D T A 3 XT" DTPA a La 13.3 8.7 12.1 15.9 Nd 14.2 9.8 13.7 18.1 Sm 14.5 10.7 14.7 18.8 Gd 14.9 11.0 14.8 18.8 Ho 16.5 12.5 14.9 19.2 Yb 17.1 12.9 14.6 19.1 0.10M[(CH 3) 4N]C1,3 3 b p. = 0.16 M NaCl 90 La Nd Sm Gd Ho Yb Ln Figure 5.4. Trends in pM values (See Table 5.4 for Conditions). Although their origins are open to much debate, stability trends for lanthanides with linear polyaminocarboxylate ligands fall into two types:31'35 1) the formation constants increase significantly with increasing charge to ionic radius ratios as found in EDTA-type ligands; 2) the stability constants increase, reache a plateau, then slightly decline, as with DTP A. The Ln-XT system follows the second trend (Table 5.4, Figure 5.4). Compared to EDTA, X T binds more selectively to early lanthanides, similarly to middle ones and less strongly to the late ones. T M D T A , however, is a weaker ligand for Ln than X T . The fact that E D T A and T M D T A have similar trends in pM across the series but XT-Ln shows a different binding profile demonstrates that the phosphinate in X T is significant. 91 5.4 Conclusions The chemistry of a series of [Ln(XT)]2", where X T is an EDTA-like ligand containing an extra phosphinate group has been compared with that of Co1 1- and Co I H -XT. The phosphinate group is not coordinated in the Co complexes but is bound in the lanthanide complexes. Solid state and solution behaviours of Ln-XT species are consistent: both mono-protonated and non-protonated species have been found. Protonation of the metal complex does not lead to dissociation of a carboxylate, rather, the proton distributes around the molecular ion. The pM values of Ln-XT are comparable to those of Ln-EDTA, but are higher than those of Ln-TMDTA. 5.5 References 1) Norman, T. J.; Parker, D.; Royle, L.; Harrison, A.; Antoniw, P.; King, D. J. J. Chem. Soc. Chem. Comm. 1995, 1877. 2) Order, S. E.; Klein, J. L.; Leicher, P. K.; Frinke, J.; Collo, C ; Carlo, D. J. Int. J. Radial Oncol. Biol. Phys. 1986,12, 277. 3) Lauffer, R. B. Chem. Rev. 1987, 87, 901. 4) Caravan, P.; Ellison, J. J.; McMurry, T. J.; Lauffer, R. B. Chem. Rev. 1999, 99, 2293. 5) Volkert, W. A.; Hoffman, T. J. Chem. Rev. 1999, 99, 2269. 6) Heeg, M . J.; Jurisson, S. S. Acc. Chem. Res. 1999, 32, 1053. 7) Berg, D. J.; Rettig, S. J.; Orvig, C. J. Am. Chem. Soc. 1991,113, 2528. 8) Yang, L.-W.; Liu, S.; Rettig, S. J.; Orvig, C. Inorg. Chem. 1995, 34, 4921. 9) Yang, L.-W.; Liu, S.; Wong, E.; Rettig, S. J.; Orvig, C. Inorg. Chem. 1995, 34, 2164. 92 10) Lowe, M . P.; Caravan, P.; Rettig, S. J.; Orvig, C. Inorg. Chem. 1998, 37, 1637. 11) Liu, S.; Yang, L.-W.; Rettig, S. J.; Orvig, C. Inorg. Chem. 1993, 32, 2773. 12) Caravan, P.; Mehrkhodavandi, P.; Orvig, C. Inorg. Chem. 1997, 36, 1316. 13) Caravan, P.; Lowe, M . P.; Read, P. W.; Rettig, S. J.; Yang, L.-W.; Orvig, C. J. Alloys Compd. 1997, 249, 49. 14) Maier, L.; Smith, M . J. Phosphorus and Sulfur 1980, 8, 67. 15) Varga, T. R. Synth. Commun. 1997, 27, 2899. 16) Huskens, J.; Sherry, A. D. J. Am. Chem. Soc. 1996,118,4396. 17) Pulukkody, K. P.; Norman, T. J.; Parker, D.; Royle, L.; Broan, C. J. J. Chem. Soc, Perkin Trans. 2 1993, 605. 18) Martell, A. E.; Hancock, R. D. Metal Complexes in Aqueous Solutions; Plenum Press: New York, 1996. 19) Glasoe, P. K.; Long, F. A. J. Phys. Chem. 1960, 64, 188. 20) Gran, G. Acta Chem. Scand. 1950, 4, 559. 21) Gans, P.; Sabatini, A.; Vacca, A. J. Chem. Soc. Dalton Trans. 1985, 1196. 22) Gomez-Romero, P.; Jameson, G. B.; Casan-Pastor, N.; Coronado, E.; Beltran, D. Inorg. Chem. 1986,25,3171. 23) McCandlish, E. F. K.; Michael, T. K.; Neal, J. A.; Lingafelter, E. C ; Rose, N. J. Inorg. Chem. 1978,i 7,1383. 24) Okazaki, H.; Tomioka, K.; Yoneda, H. Inorg. Chim. Acta. 1983, 74, 169. 25) Kennard, C. H. L. Inorg. Chim. Acta 1967,1, 347. 26) Kolat, R. S.; Powell, J. E. Inorg. Chem. 1962,1, 485. 27) Garner, C. D.; Collison, D.; Mabbs, F. E. Metal Ions Biol. Syst. 1995, 31, 631. 93 28) Sakagami, N.; Yamada, Y.; Konno, Y.; Okamoto, K. Inorg. Chim. Acta 1999, 288, 7. 29) Setyawati, I. A.; Liu, S.; Rettig, S. J.; Orvig, C. Inorg. Chem. 2000, 39, 496. 30) Varga, T. R.; Kiraly, R.; Brucher, E. Ach-Models in Chem. 1999,136, 431. 31) Kumar, K.; Jin, T.; Wang, X.; Desreux, J. F.; Tweedle, M . F. Inorg. Chem. 1994, 33, 3823. 32) Rajan, K. S.; Mainer, S.; Rajan, N. L.; Davis, J. M . J. Inorg. Biochem. 1981,14, 339. 33) Martell, A. E.; Smith, R. M . Critical Stability Constants; Plenum: New York, 1974-1989; Vol. 1-6. 34) Paul-Roth, C ; Raymond, K. N. Inorg. Chem. 1994, 34, 1408. 35) Tse, P. K.; Powell, J. E. Inorg. Chem. 1985, 24, 2727. 94 Chapter 6. Conclusions and Further Thoughts 6.1 Rhenium chemistry 6.1.1 Conclusions In the first three chapters I have discussed some Re complexes and their syntheses. Chapter 2 details the chemistry of [RevO(ppme)X]; a halide mixture of [ReO(ppme)X] (X = BrojClo.7) has been synthesized via a substitution route and structurally characterized. Alternatively, [ReO(ppme)Cl] has been synthesized from [NH4][Re04] via simultaneous reduction/complexation. Substitution of the chloride with thiocyanate or aromatic thiolates was successful; however, the one-pot synthesis of the respective ternary thiolate/ppme complexes was not. Chapter 3 describes a structural motif in Re v=0 complexes. Oxorhenium(V) complexes of H2pmen and H2bped have been synthesized and structurally characterized. Along with [ReO(bbpen)]+ and [ReO(qtp)(OCH3)2]+, the [ReO(H2pmen)Cl2]+ and [ReO(bped)]+ cations share a rare seven-coordinate structure with a distorted pentagonal bipyramidal geometry, which represents a potentially general structural type in Re v=0 complexes. Chapter 4 investigates the chemistry of H2pmen with Re in more detail. The three cations, [ReO(H2pmen)Cl2]+, [ReO(Hpmen)Cl]+ and [Re02(H2pmen)]+, are related by hydrolysis/HCl substitution. [ReO(Hpmen)Cl][Re04] was structurally characterized and found to contain a water stable amido-Re bond. Dehydrogenation of the N-donor ligand from amine to imine with concomitant two-electron reduction of the Re center occurs readily in these systems. With suitable co-ligands such as 3-hydroxy-4-pyrones, ternary complexes [ReCl(L)(Ci4Hi4N4)][Re04] (where C i 4 H M N 4 = H2pmen-4H and L = 3-hydroxy-4-pyronate 95 anions) can be made from [NH^fReCM], H 2pmen4HCl and the pyrones in one-pot syntheses. [ReCl(ma)(Ci4Hi4N4)][Re04J was structurally characterized, and was found to contain a 7-coordinate ReU I center. The reduction/complexation route from perrhenate is usually preferred, and it was found that triphenylphosphine reduction in acidic alcohols provides a general and mild reaction condition for the synthesis of Re v=0 complexes. [ReO(ppme)Cl], [ReO(H2pmen)Cl2]Cl, [ReO(Hpmen)Cl][Re04], and [ReO(bped)] [Re04] were made by this reduction. (Although not reported here, a number of [ReOL2Cl] complexes where L = 3-hydroxy-4-pyronate or a 3-hydroxy-4-pyridinonate can also be made via this route.) This reduction/complexation also affords a better yield than do substitution reactions from [(C4H9)N][ReOBr4] (because of the hydrolytic instability of the starting salt) or ReOCl3(P(C6Hs)3)2 (because the stable nature of the starting complex precludes its reaction with a weak ligand, such as ppme2"). Another interesting reduction route for perrhenate is amine dehydrogenation. Although amine dehydrogenation is ubiquitous in metal complexes containing multidentate amine ligands, its use to reduce high oxidation state metal ions is not. The ramification of such is a clean synthetic pathway to intermediate oxidation state metal imine complexes from permetallate, a route that is not always accessible. Given a reasonable inertness/stability of the linear or macrocyclic amine/imine complexes with intermediate oxidation state Tc/Re, this reaction has implications in nuclear medicine. Synthesis of a ternary complex via the reduction/complexation route is difficult, and finding a right combination of ligands is crucial. The failure of the one-pot synthesis of [ReO(ppme)SR] is a good demonstration of this. It should not be generalized, however, to 96 all "4+1" combinations of N2O2 ligands and thiolates, because ppme is a particularly weak ligand for Re and aromatic thiolates form very stable [ReO(SR)4]" species. When a good combination is found, the reaction works well. The N 4 and 3-hydroxy-4-pyrone is an interesting combination: the imine/pyridine N 4 ligand stabilizes the Re111 center and the pyronate and the chloride satisfy its charge demand. 6.1.2 F u r t h e r T h o u g h t s Re111 complexes of bped, bbpen, and a number of tripodal amine phenol ligands can probably be prepared by reducing the ReO complexes with an oxo transfer reagent in the presence of a monodentate coligand. The availability of a suitable monodentate co-ligand might be important in driving this reaction because of the preference of Re111 to be seven-coordinate and/or the need for a 71-accepter to lower the oxidation state. There are not many Re I V complexes with N, O or S donor ligands useful for nuclear medicine.1"4 Re I V is inert and, conceivably, the complexes would not be demetallated by substitution on a reasonable time-scale. Re I V ternary complexes would benefit from this inertia. The problem might lie in the synthesis; correlation of the synthesis with the tracer scale chemistry may not be possible, and reduction to Re I V could be difficult. Complexes with u.-0 ligands are prevalent on the synthetic scale,5'6 as in the case ofBa 2[Re 2 l v(u-0)2(EDTA) 2] which was prepared by reducing Re vO-(EDTA) species prepared in situ,1 but in practice the formation of di- or poly-nuclear species is highly disfavoured on the nuclear medicine scale (10"6 -10"8 M) . 8 With appropriate ligand or ligand combinations, mononuclear species can be easily prepared. A number of R e ' ^ C b (L = bidentate ligands) 97 and Re CbOSbCb) (L= Schiff base imine phenol ligands) complexes have been made by substitution.5 The ligands in this thesis, bped2', bbpen2", Tkpmen, and the oxidized versions of fkpmen (i.e. amine/imine or diimine dipyridine N 4 ligand), should be investigated in their complexation of Re I V , and reduction of their ReO complexes (or oxidation from the proposed Re111 complexes, should they be made) will prove interesting. (Reduction of [ReO(ppme)Cl] was attempted but without success.) A challenge in making the Re I V complexes would be, once again, reduction from perrhenate. With a reasonably established triphenylphosphine reduction to Re v O, a subsequent in situ one-electron reduction by another reducing agent could be useful. As more data on metal ion-induced ligand modification in Re complexes accumulates, the ligand design for Re complexes should intentionally include such features. Because Re is difficult to reduce and the ligand modifications usually involve a concomitant redox process on the metal ion, this approach would open synthetic routes to otherwise difficult oxidation states and complexes. A good example of this is the use of a dehydrogenation reaction to reduce [Re04]", as in hydrazinopyridine9 and Hipmen, forming N coordinated complexes. Also, Re has been investigated extensively as an oxo-transfer catalyst, the chemistry of which should be useful in ligand design for nuclear medicine. 1 0 - 1 2 A logical extension of Chapter 4 would be to examine the reaction chemistry of [ReCl(ma)(Ci4Hi4N4)][Re04] (4.5). Interaction of the ternary complexes with endogeneous ligands and the biochemistry of the complexes will be interesting: what will their reactive metal sites do in living systems? Preliminary studies with 4.5 showed that the reactions are curious. The complex gave an intensly blue aqueous solution (green in methanol), and the ' H N M R spectrum in D 2 0 showed that both ligands are still coordinated to the metal center 98 but in a drastically different fashion from the coordination in CD3OD. [ReCl(ma)(Ci4Hi4N4)][Re04] also reacts with cysteine and lysine in methanol, leading to a number of products, most likely ternary complexes with the amino acids. The ligand exchange is slow and cysteine did not demetallate the complex. This system should also be investigated towards its application in a "pre-targeting" drug design. Recently a bifunctional monoclonal antibody that targets tumor tissue and binds a receptor molecule has been developed.13'14 The MAb binds to the tumor when administered. At peak uptake, a chase molecule is given to clear out the MAbs not bound to the tumor, followed by a radiolabeled receptor ligand which seeks the tumor bound MAbs and tags them. This approach eliminates the problem associated with using short-lived radionuclides to monitor slow MAb-antigen reactions. From a coordination chemistry point of view, ternary complexes offer a unique possibility of such a reaction on the metal site: should the MAb bear a ligand suitable for a specific metal complex, it would most likely sequester the complex. Re/Tc complexes are especially relevant in this regard because their coordination is different from that of native metals in biological systems. Thus one could tag a ligand that binds Tc/Re selectively. The complexes described in Chapter 4 should be interesting in this regard. More generally, the exploration of Re (and Tc) reactivity in biological systems has a conceptual appeal. The in vivo setting of a biological system is not familiar with the reactions of Re. The closest biologically significant metal, in terms of ligand binding, is Cu, which is still quite different from Re. This provides a window of opportunity for exploring metal-based chemistry in living systems. With a body of developed separation/analytical 99 methods in biology/biochemistry and with the radioactivity of the metal, such studies should be feasible. The use of Re in medicine need not to be limited to exploring its nuclear properties. Re is one of the most rare elements but it is only moderately expensive. (Na[Re04] is sold for Cdn $70.50/5 g from Aldrich.) Although toxicity data are scarce, Re complexes (perrhenate and Re111 chloro complexes) are well-tolerated in animal models.15 It was found that perrhenate preferentially localizes in thyroid prior to being secreted, and even this bio-concentration is reversible.16'17 Considering the burgeoning accumulation of biological data on Tc complexes and current understanding of Re chemistry, it is surprising that there is so little study on the biochemistry of Re complexes. After all, the complexes are expected to be oxidized to perrhenate, the clearance of which is straightforward. A superior ligand combination for Re is that of P and S donor ligands. Both form stable complexes on their own or with additional ligands. Complexes with combinations of monodentate P or S ligands are common, stable and easy to make.5 With an added advantage that the ligands themselves are reducing, these combinations should attract more attention in the field of nuclear medicine. A coordination chemist would appreciate the reaction in equation 6.1: ReL + [NH 4][TcV I 10 4] > TcL + [NH4][ReVII04] [6.1] This exchange/redox reaction explores the difference between the 4d and 5d transition series: 5d metals are more easily oxidized, hence their complexes can act as reductants for pertechnetate, and this reaction is thermodynically favoured. The reality may never be this 100 simple and clean, and further chemistry of both ReL and TcL may complicate the picture, but it promises a wealth of interesting chemistry, and possibly an easy route to make Tc complexes. For the sake of simplicity, [ReO(bped)]+ systems would be a starting point for this study. 6.2 Lanthanide chemistry 6.2.1 Conclusions Chapter 5 probes the chemistry of Ln with XT, specifically, the role of an extra phosphinate in the coordination behaviour of an EDTA-like ligand. The phosphinate coordinates to the lanthanides, but not to Co 1 1 or C o m in their respective complexes. It contributes to the overall stability of the ligand, but not by much. It has a more profound effect on stability trends in the lanthanide series than on the stabilities of the individual complexes: the phosphinate eases the selectivity of an EDTA-type ligand to late lanthanides. Solid state and solution behaviours of Ln-XT species are consistent: both mono-protonated and non-protonated species have been found. Protonation of the metal complex does not lead to dissociation of a carboxylate, rather, the proton distributes around the molecular ion. 6.2.2 Further Thoughts An N M R study of the aqueous solution behaviour of XT, and to a lesser extent, Ln-X T is impeded by the high symmetry of the ligand: without suitable markers, how proton 101 loss affects the solution structure of the ligand and the complexes remaines unanswered. Thus asymmetric ligands adopting the general structure of RiR2NCH2P(0)(OH)CH2NR3R4 (Scheme 6.1) should be useful in clarifying the solution structure of the ligand and the complexes at different pHs. H 2 0 H 3 P 0 2 + CH 2 Q + R 1 R 2 N H ^ H 0 H H N R 3 R 4 / C H 2 0 6M HCI Scheme 6.1 The stepwise addition of secondary amine units to a phosphinic acid can be proposed (Scheme 6.1). A mild reaction condition forms a a-aminophosphinic acid, and a subsequent Moedritzer Irani reaction under forcing conditions will introduce the other P-C-N bond. Both reactions are known 1 8' 1 9 and ( H O O C C ^ N C ^ P C ^ F k has been prepared. The development of this general method for making RiR2NCH2P(OOH)CH2NR3R4 also has the advantage of allowing free manipulation of charge, lipophilicity, or possible conjugation of the ligand. HOOC—v / \ /—COOH ^ O H H O ^ ,N N HOOC ^ C Q O H HOOd C ° ° H Scheme 6.2 102 The step-wise Moedritzer Irani reactions also allow rapid synthesis of a variety of large ligands, for example, a N4O8 donor set ligand (Scheme 6.2) can be prepared via the proposed two-step-synthesis from readily available starting materials. This ligand is interesting in its own right. It conceivably can coordinate to two metals ions and the system promises some rich chemistry, for instance, interactions of the metals and cooperativity in metal ion binding. Variations on the ligand structure and donor types could be introduced easily to further the understanding of these systems. 103 6.3 References 1) Technetium, Rhenium and Other Metals in Chemistry and Nuclear Medicine; S. G. Editorali: Padova, 1999; Vol. 5. 2) Clarke, M . ; Podbielski, L. Coord. Chem. Rev. 1987, 78, 253. 3) Deutsch, E. ; Libson, K.; Jurisson, S.; Lindoy, L. F. Prog. Inorg. Chem. 1983, 30, 75. 4) Dilworth, J. R.; Parrott, S. J. Chem. Soc. Rev. 1998, 27, 43. 5) Connor, K. A.; Walton, R. Comprehensive Coordination Chemistry; Wilkinson, G., Gillard, R. D. and McCleverty, J. A., Eds.; Pergamon Press: Oxford, 1987; Vol. 4. 6) Rouschias, G. Chem. Rev. 1974, 74, 531. 7) Ikari, S.; Ito, T.; McFarlane, W.; Nasreldin, M . ; Ooi, B.-L.; Sasaki, Y.; Sykes, A. G. J. Chem. Soc. Dalton Trans. 1993, 2621. 8) Liu, S.; Edwards, D. S. Chem. Rev. 1999, 99, 2235. 9) Hirsch-Kuchma, M . ; Nicholson, T.; Davison, A.; Davis, W. M . ; Jones, A. G. Inorg. Chem. 1997, 36, 3237. 10) Brown, G. M ; Weaver, T. R.; Keene, F. R.; Meyer, T. J. Inorg. Chem. 1976,15, 190. 11) Brown, S. N.; Mayer, J. M . Inorg. Chem. 1992, 31, 4091. 12) Abu-Omar, M . M . ; Khan, S. I. Inorg. Chem. 1998, 4979. 13) Volkert, W. A.; Hoffman, T. J. Chem. Rev. 1999, 99, 2269. 14) Heeg, M . J.; Jurisson, S. S. Acc. Chem. Res. 1999, 32, 1053. 15) Jacquet, P. Handbook on Toxicity of Inorganic Compounds; Seiler, H. G., Sigel, H., and Seigel, A. , Eds.; Marcel Dekker Inc.: New York, 1988. 104 16) Lin, W. Y.; Hsieh, J. F.; Tsai, S. C ; Yen, T. C ; Wang, S. J.; Knapp, F. F. Nucl. Med. Biol. 2000, 27, 83. 17) Kotzerke, J.; Fenchel, S.; Guhlmann, A.; Stabin, M. ; Rentschler, M . ; Knapp, F. F.; Reske, S. N. Nucl. Med. Comm. 1998,19, 795. 18) Bazakas, K.; Lukes, I. J. Chem. Soc. Dalton Trans. 1995, 1133 and references therein. 19) Moedritzer, K.;Trani, R. R. J. Org. Chem. 1966, 31, 1603. 105 Appendix Selected Crystallographic Data for the X-ray Crystal Structures Reported in This Thesis Table A l . Selected Crystallographic Data for [ReO(ppme)X] X = Br0.29/Clo.7i. formula fw crystal system space group a, A c, A V, A 3 Z Pcalc, g/cm3 F(000) temperature radiation u,, cm"1 transmission factors 26max total reflections unique reflections reflections with I > 3CT(I) no. of variables R; R w C18H24Bro.29Clo.7iN205P2Re 644.90 tetragonal P4i 12.898(1) 13.456(2) 2238.4(5) 4 1.913 1252.88 21.0°C Mo-Ka(X=0.71069 A) 62.09 0.653- 1.000 6 5 . 0 ° 4517 4207 2489 272 0.032; 0.026 R = Z I |F 0 | - |FC| I /2 |F 0 | , I > 3a(I); R w = (ZW(|F 0| - |FC|) 2/ZW(F 0 2))' / 2 106 CD X t O U O > Pi <D O O U • fl > a ts Q o & OB _o ~5 +-» co U T3 ( D +-» O o 13 H C a, X) o Pi X O u o cu 0i T 3 U CM o o u C N u U s o t o o Oil OH m o (J-H oo Os Os oi oi o cn u u CN 00 d rr 00 < SO o '3 o c o 3 o 1 o •s o o o |3 "o o e o 3 CD 1/5 O o OH oT~ oT oT r f «—i o CN CN oo CN o CN r t ON ,_; o CN CO o cn CN r r SO r--od so i n CN cn CN r ^ so' Os SO SO rr oo oo , — N rl- oT 00 C N , no C N ' Os SO Os C N cn rT r f C N cn © d o C N Os C N CN OH O JH 00 § o < °< »< -a °< 00 U O O O CN >-H CN £ Os © cn © CN — ' i o Os • o ' o o a o o o o c o 1 o o o o ] cn 1 r- OS SO o i n CN oo o r—1 rr, cn m rn rr so o o d d d d cn o Y - H d cn o d cn CN rt" o CN l> CN SO rl-t—1 oo d r f Os so o i> (N cn SO /• \ r r o O (N d d d o o o o o o 00 rt-CN o C M 3 o 5L cn <—i m r--C N m o o d d I rN w I Table A3. Selected Crystallographic Data for 4.2 and 4.5. complex 4.2 4.5 empirical formula C u H n ^ O s R ^ C l C2iH23N 4 0 8Re2Cl fw 729.18 867.29 colour, habit green, platelet red-brown, block crystal system triclinic monoclinic space group P\ Plxln a, A 9.3975(7) 8.4833(2) b,A 10.648(2) 18.7387(7) c, A 10.779(2) 15.0656(6) adeg 68.78(1) ytfdeg 74.13(1) 91.47(1) r d e g 68.36(1) V, A 3 922.3(2) 2394.1(1) z 2 4 Pcalc, g/cm3 2.626 2.403 radiation Mo-Koc Mo-Ka ix, mm"1 13.30 10.28 transmission factors 0.466- 1.000 0.684- 1.000 R;Rw 0.044; 0.118 0.027; 0.072 R = S I |F 0 | - |FC| I /Z |F 0 | ,1 > 3a(I); Rw = (E(F 0 2 - F c 2) 2/Sw(F 0 2) 2) 108 Table A4. Selected Crystallographic Data for K 3[Co(XT)]-2H 20. formula fw crystal system space group a, A b, k c, A v, A 3 z Pcalc, g/cm3 F(000) temperature radiation u., cm"1 transmission factors 20max total reflections unique reflections observations with I > 3a(I) no. of variables R;Rw C 1 0 H 1 6 C o K 3 N 2 O 1 2 P 563.44 orthorhombic P2j2i2 11.8931(5) 8.4428(11) 9.6761(4) 971.6(2) 2 1.926 570.00 - 9 3 ± 1 ° C Mo-Ka (A,=0.71069 A ) 16.74 0.7308- 1.0000 60.1° 8942 1603 2002 165 0.036; 0.040 R = S I |F 0 2 | - |F C 2 | | /I|F ( 2 | ,I>3a(I); Rw = (l(F 0 2 - F c 2 ) 2 /Iw(F 0 2)T 109 

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.831.1-0061239/manifest

Comment

Related Items