The Cyclotron Production and Cyclometalation Chemistry of 192-Ir

Graeme Langille

B.Sc., Simon Fraser University, 2012

Thesis Submitted in Partial Fulﬁllment of the Requirements for the Degree of

Master of Science

in the Department of Chemistry Faculty of Science

c Graeme Langille 2014 SIMON FRASER UNIVERSITY Fall 2014

All rights reserved. However, in accordance with the Copyright Act of Canada, this work may be reproduced without authorization under the conditions for ”Fair Dealing”. Therefore, limited reproduction of this work for the purposes of private study, research, criticism, review and news reporting is likely to be in accordance with the law, particularly if cited appropriately. APPROVAL

Name: Graeme Langille

Degree: Master of Science (Chemistry)

Title: The Cyclotron Production and Cyclometalation Chemistry of 192-Ir

Examining Committee: Chair: Dr. Hua-Zhong Yu Professor

Dr. Corina Andreoiu Senior Supervisor Associate Professor

Dr. Paul Schaﬀer Co-Supervisor Adjunct Professor

Dr. Tim Storr Supervisor Associate Professor

Dr. Krzysztof Starosta Supervisor Associate Professor

Dr. Robert Young Internal Examiner Professor

Date Defended/Approved: December 11, 2014

ii Partial Copyright Licence

iii Abstract

The goal of this thesis is to demonstrate the cyclotron production, radiochemical isolation, and cyclometalate chemistry of radio-iridium isotopes. In recent work, Luminescence Cell Imaging (LCI) has been combined with radioisotopes, leading to compounds that can be imaged with both optical microscopy and nuclear techniques. Radiometals excel in this multifunctional setting, providing ideal chemical and nuclear properties for luminescence, biological targeting, nuclear diagnostics, and therapy. Iridium cyclometalate compounds have demonstrated potential in LCI with excellent photophysical properties. Independently, low specific activity 192Ir has been successfully applied in brachytherapy as a high-intensity β− emitter. Despite this, radio-iridium has not yet been applied to cyclometalate chemistry, nor a radiochemical isolation method developed for its cyclotron production. Herein is demonstrated the feasibility of the production and isolation of radio-iridium, and its application to cyclometalate chemistry as a potential tool for nuclear medicine research. Natural osmium was electroplated onto a silver disc from basic media, and the thin deposits obtained were weighed and characterized with scanning electron microscopy. These targets were irradiated using the TRIUMF TR13 cyclotron, delivering 12.7 MeV protons to the target disc to access the AOs(p, n)AIr reaction channels. Three irradiations were performed at 5 µA for 1 hour, and one at 20 µA for 2 hours. Gamma spectra of the targets were collected and a range of iridium isotopes (186-190, 192) identified and quantified. The irradiated material was then oxidized, dissolved from the target backing, and separated via anion exchange. Once isolated, the isotopes were applied to an adapted cyclometalation procedure, and the compounds were identified and quantified against non-radioactive standards using high performance liquid chromatography with coupled γ-ray and ultraviolet detectors. The procedure developed here has enabled the study of radio-iridium cyclometalates, a potentially new class of theranostic compounds for nuclear medicine.

Keywords: nuclear medicine; radiochemistry; radiopharmacy; radiometals; theranostics; radiochemical separations; isotope production; 192Ir; cyclometalate chemistry

iv Acknowledgments

The breadth of this project required the support and direction of a diverse group of people, without whom I would most certainly still be toiling away in the lab. It has been a privilege to work with Dr. Corina Andreoiu, whose enthusiasm for nuclear science inspired my own exploration. I am grateful to Dr. Paul Schaffer for sharing his broad knowledge of nuclear medicine, and constant positive attitude. My committee members, Dr. Tim Storr and Dr. Krzysztof Starosta, have my gratitude for not only their input in this project, but also their ongoing roles throughout my scientific education. There are many members of the Nuclear Medicine Division at TRIUMF to whom I am indebted: for his targetry prowess and laboratory conversations, Dr. Stefan Zeisler; for their inorganic expertise, Dr. Hua Yang and Dr. Qing Miao; and, for their expert handling of the cyclotron, Linda Graham and Dave Prevost. At SFU I had the fortune to work with many other skilled individuals: Dr. Jean-Claude Brodovitch, an excellent radiochemist and human being; Dr. Bob Young, for his review of my thesis and the occasional use of his lab; the members of the Storr group, for enduring my occupation of their fume hoods; and finally, the Starosta and Andreoiu groups, for their support and friendship.

v Contents

Approval ii

Partial Copyright License iii

Abstract iv

Acknowledgmentsv

Contents vi

List of Tables viii

List of Figures ix

1 Introduction1 1.1 Nuclear Medicine Basics...... 2 1.1.1 Radioactive Decay...... 2 1.1.2 Nuclear Reactions...... 4 1.1.3 Radiopharmacy: Imaging and Therapy...... 8 1.2 Iridium cyclometalate Chemistry...... 10 1.2.1 Luminescence Cell Imaging and Transition Metal Lumophores... 10 1.2.2 Iridium cyclometalates...... 12 1.2.3 Radio-Iridium...... 14 1.3 Thesis Overview...... 15

2 Targetry 16 2.1 Experimental...... 16 2.1.1 Reagents and Instrumentation...... 16

2.1.2 Distillation of OsO4 ...... 17 2.1.3 Osmium Electroplating...... 17 2.2 Discussion...... 18

vi 2.2.1 Proton Attenuation in Matter...... 20 2.3 Conclusion...... 20

3 Isotope Production 21 3.1 Experimental...... 22 3.1.1 Osmium Target Bombardment...... 22 3.1.2 Gamma-Ray Spectroscopy...... 23 3.1.3 Isolation of Radio-Iridium...... 24 3.2 Discussion...... 25 3.2.1 Irradiation...... 25 3.2.2 Isotope Identiﬁcation and Quantiﬁcation...... 26 3.2.3 Radiochemical Separation...... 29 3.3 Summary...... 30

4 Cyclometalation 31 4.1 Experimental...... 31 4.1.1 Reagents & Instrumentation...... 31 4.1.2 Non-Radioactive Syntheses...... 32 4.1.3 Carrier-Added Radio-Synthesis of Compound 2 ...... 33 4.2 Discussion...... 34 4.2.1 Radiosynthesis...... 36 4.3 Summary...... 37

5 Conclusion 38

Bibliography 40

Appendix A Derivations and Characterization Data 45

vii List of Tables

1.1 Types of radioactive decay...... 2

3.1 Isotopic composition of naturally abundant osmium...... 21 3.2 Produced nuclides and characteristic γ-rays...... 27 3.3 Calculated isotope yields...... 28

4.1 HPLC conditions for reaction component separation...... 34

viii List of Figures

1.1 An excerpt from the chart of nuclides...... 3 1.2 192Ir decay product level schemes...... 5 1.3 Cyclotron design schematic...... 6 1.4 192Os(p, n)192Ir reaction excitation function...... 7 1.5 Radiation interactions with biological matter...... 8 1.6 Operating principle of targeted radiopharmacy...... 9 1.7 Three generations of diagnostic radiopharmaceutical...... 10 1.8 Bimodal imaging agents...... 11 1.9 Cyclometallate luminescence cell imaging example...... 12 1.10 Two step cyclometalation reaction scheme...... 13 1.11 Representative iridium cyclometalate compounds...... 13 1.12 A schematic pathway of the cyclometalation reaction...... 14 1.13 The chart of nuclides focused on long lived iridium isotopes...... 15

2.1 Osmium target electroplating methodology...... 18 2.2 Scanning electron micrographs of the osmium target material...... 19

3.1 Polymer structure of Dowex 1X8 anion exchange material...... 22 3.2 The TRIUMF TR13 cyclotron...... 23 3.3 Target holder employed in all irradiations...... 24 3.4 A sample spectrum collected from an irradiated target plate...... 25 3.5 Elution proﬁle of radio-iridium from anion exchange column...... 26 3.6 The radiochemical separation procedure...... 29

4.1 Reaction scheme for the synthesis of the dimer compound 1 ...... 32 4.2 Reaction scheme for the synthesis of 2 ...... 33 4.4 UV absorption calibration curve...... 35 4.5 UV-HPLC traces of reaction components...... 36 4.6 Overlaid radio- and UV-chromatograms of compounds 1 and 2 ...... 37

ix Chapter 1

Introduction

Radioactivity has played a fundamental role in the development of modern medicine. The radiation produced by the different types of nuclear decay have applications unmatched by conventional diagnostic and therapeutic methods. Furthermore, the observational power afforded by the tracer principle has led to many discoveries fundamental to our current understanding of human physiology and biochemistry. As these systems are better understood, however, new questions arise that demand new methods be explored. Central to the study and practice of nuclear medicine is the radiopharmaceutical, a drug that incorporates a radioactive atom. Implicit in this composition is the breadth of science it encompasses, borrowing heavily from various disciplines within physics, chemistry, and biology. The successful development of a radiopharmaceutical requires careful consideration from each perspective, and an understanding of the limitations each imposes on the others. The field of molecular imaging uses several imaging modalities to observe the interactions of target molecules in their associated biological processes in vivo [1,2,3]. Its rapid progression has created a demand for new agents that may be imaged from different perspectives, introducing the concept of multifunctional radiopharmacy. In multimodal imaging, the interactions of a drug with the system may be observed at different scales [4,5], while so-called ‘theranostic’ agents combine an imaging modality with a therapeutic function to allow real-time observation of the treatment efficacy [6]. Limiting the development of these drugs are the isotopes currently available from nuclear medicine, which lack the required chemical and nuclear properties. Transition metals exhibit versatile redox chemistry that can be adapted to many applications, and their isotopes encompass a broad range of nuclear decay properties [7,8]. These radiometals are of interest to Luminescence Cell Imaging (LCI) [9], a modality that has been incorporated into several multifunctional drugs [10, 11, 12, 13]. Iridium cyclometalate compounds are uniquely suited to this application, exhibiting excellent stability and

1 luminescence properties [14, 15, 16]. Despite this potential, however, the radiopharmaceutical chemistry of iridium has not yet been explored. Herein is described a comprehensive methodology for the radiolabelling of a highly luminescent iridium compound, including its cyclotron production and radiochemical isolation.

1.1 Nuclear Medicine Basics

1.1.1 Radioactive Decay

A nucleus is a group of protons and neutrons, collectively named nucleons, that are bound together by the strong nuclear force. As in the electronic structure of an atom, there are conﬁgurations of nucleons that are inherently unstable. Atoms may resolve their instability through reaction with other species, or deactivation to a lower energy; nuclei, however, are generally shielded from interaction with each other by their electronic orbitals, and have only deactivation available. This deactivation may take several forms, and forms the basis of radioactive decay. The most common decays, along with their use in nuclear medicine and a representative isotope, are summarized in Table 1.1, and are represented by diﬀerent colours in Figure 1.1.

Table 1.1: A summary of the most common types of radioactive decay, including their use in nuclear medicine, and the decay properties of a representative isotope. Energy values reported for β+/− decays represent the average spectral energy [17].

Decay NM Application Sample isotope t1/2 Erad (keV) α Localized Therapy 211At 7.2 h 5870 225Ac 10.0 d 5830 β+ Imaging 11C 20.3 m 386 18F 110 m 250 89Zr 78.4 h 396 β− Localized Therapy 192Ir 73.8 d 179 90Y 64 h 934 131I 8.0 d 182 γ (IT, etc.) Imaging 99mTc 6.0 h 141 111In 2.8 d 171/245 Spontaneous - 252Cf 2.65 y - Fission

2 Decay Characteristics

The activity A is equal to the number of decayed nuclides dN per unit time dt: dN A = − . (1.1) dt Solving this equation yields the radioactive decay law, which speciﬁes the number of nuclides

N at time t, given an initial population N0 and decay constant λ:

−λt N = N0e . (1.2)

Solving equation (1.2) for t and setting the fraction N = 1 gives the amount of time N0 2 required for half the population to decay, or the half-life t1/2. This quantity is measured experimentally and may be used to determine the decay constant, a quantity unique to each radionuclide: ln 2 λ = . (1.3) t1/2

16Ne! 17Ne! 18Ne! 19Ne! 20Ne! 21Ne! 22Ne! 23Ne! 24Ne! 25Ne! 26Ne! 27Ne! 28Ne! 29Ne! 30Ne! 31Ne! 32Ne! 33Ne! 34Ne! 14F! 15F! 16F! 17F! 18F! 19F! 20F! 21F! 22F! 23F! 24F! 25F! 26F! 27F! 28F! 29F! 30F! 31F! 12O! 13O! 14O! 15O! 16O! 17O! 18O! 19O! 20O! 21O! 22O! 23O! 24O! 25O! 26O! 27O! 28O! 10N! 11N! 12N! 13N! 14N! 15N! 16N! 17N! 18N! 19N! 20N! 21N! 22N! 23N! 24N! 25N! 8C! 9C! 10C! 11C! 12C! 13C! 14C! 15C! 16C! 17C! 18C! 19C! 20C! 21C! 22C! 23C! Decay Mode! 6B! 7B! 8B! 9B! 10B! 11B! 12B! 13B! 14B! 15B! 16B! 17B! 18B! 19B! 20B! 21B! EC/β+! 5Be! 6Be! 7Be! 8Be! 9Be! 10Be! 11Be! 12Be! 13Be! 14Be! 15Be! 16Be! -

Protons ! β 3Li! 4Li! 5Li! 6Li! 7Li! 8Li! 9Li! 10Li! 11Li! 12Li! 13Li! α 3He! 4He! 5He! 6He! 7He! 8He! 9He! 10He! p! 1H! 2H! 3H! 4H! 5H! 6H! 7H! n! n! stable!

Neutrons!

Figure 1.1: An excerpt from the chart of nuclides demonstrating the line of stable elements, and the general arrangement of radioactive decays around it.

β-Decay

The decays observed in this project are β− and electron capture (EC), which are classiﬁed along with β+ as beta decay. The chart of nuclides, a small portion of which is shown in Figure 1.1, qualitatively aids understanding of these and other nuclear phenomena by placing nuclides on Cartesian axes according to their proton and neutron number. Nuclides located below the line of stable elements are classiﬁed as neutron rich, and β− decay with the conversion of a neutron to a proton. To satisfy the laws of conservation of charge and lepton − − number, an electron and electron anti-neutrino (e /β , andν ¯e) are emitted, respectively.

3 The opposite is true of proton rich nuclides, which β+ decay via conversion of a proton + to a neutron with the emission of a positron and electron neutrino (β , νe). Finally, in EC nuclides capture an orbital electron, and emit an electron neutrino while converting a proton to a neutron. These nuclear reactions are summarized in equations (1.4)-(1.5):

A A − ZXN → Z+1YN−1 + β +ν ¯e− , (1.4) A A + ZXN → Z−1YN+1 + β + νe− , (1.5) A − A ZXN + e → Z−1YN+1 + νe− . (1.6)

A more thorough description of these and other nuclear decay processes may be found in many texts [18]. A nuclide may decay through more than one decay mode, each with a defined probability or branching ratio. Furthermore, all radioactive decays may populate an excited state of the daughter nuclide, which deactivate by characteristic γ-ray emission. In Figure 1.2, the EC and β− decay branches of 192Ir populate several excited states of the daughter nuclides 192Os and 192Pt, respectively, with defined probabilities. If they are of high enough intensity, the deactivation γ-rays may be imaged with the techniques described in Section 1.1.3; more often they contribute to an increased patient radiation dose. For an isotope to be medically useful from a nuclear perspective, several key requirements must be met: 1) the half-life must be appropriate for the biological process under study; 2) the decay radiation must have a suitable energy, whether for imaging or therapy (discussed further in Section 1.1.3); 3) there must not be an excess of alternate decay branches, leading to excessive radiation dose; 4) the daughter nuclide must not contribute significantly to the radiation dose. Once an isotope with the proper nuclear and chemical properties is selected, attention must turn to its production.

1.1.2 Nuclear Reactions

Accelerators

Of the radionuclides naturally found on earth, few have decay properties of interest to nuclear medicine. As such, all isotopes used today are generated by nuclear reactions in an accelerator or reactor. The cyclotron is a class of particle accelerator that has seen widespread implementation in hospitals and research facilities around the globe [19]. Cyclotrons accelerate charged particles in a circular trajectory using magnetic and oscillating electromagnetic ﬁelds. Arising from a comparison of the magnetic component of the Lorentz force (FB) and the centripetal force (Fc), equations (1.7) and (1.8) describe the

4 192Ir – 73.8 d! 4+! 0.0 keV!

- EC: 4.76 %! β : 95.24 %! I (%)! 3-! 177! 594! 766! 1062! 1378! 1378 keV! 0.103!

4+! 280! 417! 589! 885! 1201 keV! 5.60!

I (%)! 4+! 3+! 329! 421! 704! 910 keV! 0.094! 136!309! 604! 921 keV! 41.42! 4+! 468! 785 keV! 47.98! 3+! 690 keV! 3.93! 110! 201! 484! 2+! 4+! 296! 613! 613 keV! 375! 580 keV! 0.670! 2+! 283! 489! 489 keV!

2+! 317! 317 keV! 2+! 206! 206 keV!

0+! 0.0 keV! 0+! 0.0 keV! 192Os – Stable! 192Pt – Stable!

Figure 1.2: Electron capture and β− decay of 192Ir, demonstrating the levels, branching ratios, and deactivation energies into the daughter nuclides 192Os and 192Pt, respectively [17].

force equation and orbit frequency of a particle in a cyclotron, respectively:

mv2 F = qvB = = F , (1.7) B R c v qB ω = = , (1.8) orb 2πR 2πm where q is the charge of the particle, v is its tangential velocity, B is the magnetic field, m is the mass of the particle, R is the radius of orbit, and ωorb is the rotational frequency of orbit. If charge, mass, and field are held constant, then the ratio of velocity to radius, as well as ωorb, are also constant. This is significant in that particles at any radius have the same frequency of rotation, which allows them to be accelerated simultaneously in a continuous beam. To accelerate the particle, an oscillating electromagnetic field is applied to two or more electrodes that sandwich the particle trajectory, as in Figure 1.3. Solving equation (1.7) for v, and substituting into the kinetic energy EK reveals the dependence of the particle’s energy on its radius in the cyclotron: 1 1 E = mv2 = q2B2R2. (1.9) K 2 2m

5 Electromagnet!

Beam! - Injection! H ! Stripping Foil!

H+! !

±! ± Radio Frequency Electromagnetic Oscillator!

Figure 1.3: Schematic of the original cyclotron design, demonstrating the magnetic and electromagnetic ﬁelds employed to accelerate particles in curved trajectories.

Most modern cyclotrons accelerate hydrogen and deuterium as negatively charged ions to ease extraction. By placing a light foil (carbon or beryllium) at the desired radius, or kinetic energy as per equation (1.9), the electrons are stripped from the ion. This reverses the charge of the ion and imparts an oppositely curved trajectory out of the cyclotron, where it may then be directed at a target for isotope production. As will be discussed shortly, a variety of particle energies may be required to satisfy diﬀerent nuclear reactions.

Reaction Kinetics and Isotope Production

There are many potential reaction channels for any given projectile and target nuclide. The probability that the particles will interact and a particular channel will progress is related to the nuclear cross section (σ) of that channel, measured in units of barns (1 b = 1 × 10−24 cm2). The main contributions to cross section are the Q-value of the reaction, the Coulombic interaction of the interacting particles, and the level density of the nuclei produced in the reaction [18]. Though cross sections may be predicted with various nuclear reaction models, they must be measured experimentally for reliable data. Excitation functions, like that shown in Figure 1.4, show the behaviour of the cross section across a range of beam energies. By integrating the cross section with respect to energy, the production yield of a thick target (with deﬁned beam entrance and exit energies) is obtained. This data is useful in selecting an optimum beam energy for isotope production. The thickness of the target plays a role in the production of isotopes, as the incident particle energy is attenuated by the material it traverses. Thick targets are those that attenuate the beam energy measurably, thereby accessing diﬀerent cross sections according

6 80! Hilgers 2005! 70! Szelecsenyi 2010! 60!

50!

40! (mb) ! σ 30!

20!

10!

0! 5! 7! 9! 11! 13! 15! 17! 19! Energy (MeV)!

Figure 1.4: Excitation function for the 192Os(p, n)192Ir reaction demonstrating the magnitude of the cross section as a function of incident beam energy [20, 21].

to the excitation function. Likewise, thin targets attenuate the beam energy negligibly, and a single value of the cross section may be used to describe the interaction. The thin target production equation describes the activity A formed during isotope production:

A = λN = σnI(1 − e−λt) , (1.10) where N is the number of nuclei formed, λ is the isotope decay constant (s−1), n is the target number density (cm−2), I is the beam current (particles s−1), and t is the bombardment time (s). Rearranging equation (1.10) and making several substitutions yields an expression describing the cross section of a reaction given the activity of its product, considered by each isotope A: aAM σ = −λt , (1.11) χmNAI(1 − e ) where a is the area of the target irradiated (cm2), M is the molar weight of the isotope (g mol−1), χ is the natural abundance of the isotope, expressed as a fraction, m is the mass −1 of the target material, and NA is Avogadro’s number (mol ). A complete derivation of equations (1.10) and (1.11) is given in AppendixA.

Radiation Interactions with Matter

Understanding the behaviour of radiation in matter is essential to nuclear medicine, and will be discussed here in terms of radiation dosimetry. Dosimetry attempts to quantify the

7 energy deposited by radiation, an important consideration in both diagnosis and therapy. Each class of radiation described in Table 1.1 interacts differently with matter, biological or otherwise, dependent on its energy. For this thesis discussion will be limited to therapeutic isotopes, particularly β− emitters. A β− particle interacts with matter primarily through the Coulomb fields of the orbiting electrons and nucleus of the absorber. While collisions with nuclei are nearly elastic, energy is transferred from the β− particle to the absorber electrons, causing their excitation, ionization, and even breaking of associated bonds [22]. As demonstrated in Figure 1.5, these actions damage the genetic material of biological tissue either directly, through cleavage of DNA strands, or indirectly, through the formation of reactive chemical species. To quantify the energy deposited by radiation in different scenarios, radiation dosimetry employs several parameters: the linear energy transfer (LET), specific ionization, and range of the particle emitted [23]. Each of these quantities are dependent on the particle mass, charge, and energy, as well as the density and atomic number of the absorbing medium. The lower LET and specific ionization of β− radiation equates to lower efficacy compared with α particles, however, these properties also render β− emitters safer to work with, and they are commonly used in radiopharmaceutical and brachy- therapies [24].

β-

Repair!

Death!

H2O! Mutagenesis!

H+ + OH-! . . . H OH HO2 H2O2!

Figure 1.5: The diﬀerent actions and outcomes of radiation interactions with biological matter.

1.1.3 Radiopharmacy: Imaging and Therapy

The ability of radiation to treat and diagnose disease was recognized shortly after its dis- covery, and many later advances in nuclear science were quickly followed by applications to medicine [25]. While early experiments applied radiation externally for crude imaging and

8 therapy, its potential was realized in the development of targeted, radiolabeled pharmaceu- ticals, whose operating principle is illustrated in Figure 1.6.

Figure 1.6: Operating principle of targeted radiopharmacy, including imaging and therapy. Adapted from the Opensource Handbook of Nanoscience and Nanotechnology [26].

Parallel to nuclear techniques, a molecular understanding of disease has developed across several disciplines, leading to new generations of radiopharmaceuticals selective to various diseases and tissues. The first radiopharmaceuticals employed crude targeting that re- flected the concurrent understanding of human physiology and biochemistry, such as the 32 [ P]Na2HPO4 applied in metabolism studies by de Hevesy [27]. The biodistribution of these first generation compounds was non-specific, and they were used to measure basic physio- logical functions. Second generation radiopharmaceuticals connected the chemical structure of a compound to its biodistribution, targeting specific organs to generate perfusion imaging agents that see widespread use today. Finally, third generation radiopharmaceuticals incorporate biological targeting agents (peptides, proteins, antibodies, etc.) that interact with a receptor target, achieving specific binding in the desired tissue. Examples of popular technetium compounds representing each generation are included in Figure 1.7. As drug specificity increases, several benefits to imaging and therapy become apparent: comparable outcomes may be achieved with lower doses; the dose to the surrounding tissue decreases; and, the signal/background ratio improves, boosting contrast and image quality in imaging. Likewise, improved radiopharmaceuticals lead to a better understanding of the molecular basis of disease [28].

9 I. II. III. R O N S S R R O N C N Tc C C N N Tc Tc O C C N O N C N O R R N Cl R

Pertechnetate Sestamibi TRODAT R = methoxyisobutyl

Figure 1.7: Three generations of diagnostic radiopharmaceutical, illustrated by popular technetium compounds. Their targets are: I. thyroid (rudimentary), II. myocardial perfusion and breast tumours, and III. neuronal dopamine transporter.

Diagnostic techniques are considered to be the core of Nuclear Medicine. Their utility stems from the tracer principle, wherein minute quantities of a radiolabeled, biologically ac- tive compound are applied to a system for observation without perturbation. Positron Emis- sion Tomography (PET) and Single Photon Emission Computed Tomography (SPECT) form the basis of the study. Both techniques were instrumental in developing the current understanding of human physiology, and maintain relevance today in both research and clinical settings [25]. In consideration of the background dose absorbed by the patient, therapeutic treatments require more speciﬁc targeting. As such, the study of targeted radiotherapy has lagged behind its diagnostic counterpart, which can apply smaller amounts of radiation for good imaging. Many texts contain a further discussion of the isotopes, radiopharmaceuticals, and detectors employed in nuclear medicine [23].

1.2 Iridium cyclometalate Chemistry

1.2.1 Luminescence Cell Imaging and Transition Metal Lumophores

The high resolution of optical microscopy has found application in molecular imaging by pairing photoluminescent, targeted compounds with confocal microscope techniques to form a new imaging modality, Luminescence Cell Imaging (LCI) [29]. The images generated reveal the interactions of the imaging agent and host on a cellular level, detail that is unparal- leled by any alternate functional imaging modality; compare the sub-µm resolution of LCI with PET/SPECT (1 - 2 mm), Computed Tomography (50 µm), or Magnetic Resonance Imaging (10 - 100 µm) [1]. The photon wavelengths that enable such high resolution, however, are also quickly absorbed by tissue and currently limit practical application to in vitro

10 analysis. A class of imaging agents has emerged in response, combining LCI with nuclear techniques to yield an analysis tool capable of molecular imaging at two scales [30, 31, 32]. Figure 1.8 describes several compounds capable of imaging via both PET/SPECT and optical imaging.

a) b) S O O O H H O OH N N OH O H N N N H H H N NH2 O O O Br N N N M H O OC CO CO N N N M a) & b): M = Re, 99mTc OC CO CO

O O O c) d) S N N VEGF CdTe N H H O Cl ZnS Zr HN OH Cl COOH N N 64Cu N N HOOC COOH Figure 1.8: Examples of bimodal imaging agents. a) A lysine derived amino acid chelate, easily appended to a peptide-based targeting system [33]; b) A metal-glutamine chelate for tumour targeting [34]; c) A zirconocene compound with good luminescent properties, and potential for radiolabelling with 89Zr [35, 36]; d) A VEGF protein is coupled with a luminescent CdTe quantum dot, and a DOTA-64Cu complex appended for optical/PET imaging [37].

Though LCI may be used to image the natural fluorescence exhibited by different cell structures, termed autofluorescence, its full potential is realized in the application of a photoluminescent imaging agent, or lumophore. A sample lumophore and a representative image are shown in Figure 1.9. As in targeted radiopharmacy, these lumophores often incorporate biological vectors which localize on the desired structure before imaging. As a result, this field also benefits from advances in biochemistry and immunology, and provides them with valuable data and feedback [23]. Furthermore, some lumophores have optical properties sensitive to their chemical environment, a capacity not available through nuclear 2+ methods. Some example analytes include pH, Ca ,O2, and cAMP [32]. A detailed discussion of the confocal microscope apparatus and analysis procedures is given in [29]. Photoluminescence is defined here as electronic excitation through photon absorption, and the subsequent electronic relaxation accompanied by the emission of a photon [39].

11 O N OH N C H O N OH Ir HO N C OH N MCF-7 cell micrograph

Figure 1.9: An LCI plate of MCF-7 cells, and the iridium cyclometalate compound used to generate the image [38].

Traditional lumophores are organic compounds, and rely on delocalized electronic structures to generate short-lived fluorescence. The paired electron spins of the singlet excited state have an allowed transition to the ground state, and so excitation is short-lived. Operating on a similar timescale are the naturally fluorescent components of the cells themselves, whose autofluorescence can wash out the signal of the applied lumophore. Phosphorescence, as exhibited by many transition metal lumophores, is characterized by triplet excited states where both electrons have parallel spins. Their deactivation via photon emission is a forbidden transition, and as a result, the lifetime of these states (

1.2.2 Iridium cyclometalates

The ideal photophysical properties of iridium cyclometalate compounds have generated interest in several applications [14], including LCI [40]. These compounds excel in this application with many favourable attributes, including: 1) large Stokes’ shifts; 2) long luminescence lifetimes; 3) high quantum yields; 4) tunable emission spectra via ligand choice; 5) synthetic versatility for ease of labelling; and, 6) excellent photochemical and physicochemical sta- bilities. The Ir(III) d6 centre typically assumes an octahedral geometry defined by three metal-containing rings, two or three of which contain a metal-carbon bond. The general cyclometalate structure is shown in Figure 1.10 as Compound 2; a few select cyclometalating and ancillary ligands are shown in Figure 1.11, demonstrating structural modifications that enable different phosphorescence properties, sensing abilities, or biological targeting [40]. The photophysical performance of these compounds depends on two key structural components: the iridium core, and the cyclometalating ligands. Efficient, room-temperature phosphorescence is only observed in heavy transition metal complexes, such as iridium, as a

12 N N C H C Cl C . Δ IrCl3 3H2O + Ir Ir N C Cl C N N [1]

X N N N C Cl C N N Δ C Ir Ir + Ir C Cl C N C N N N N [2]

Figure 1.10: Two step cyclometalation reaction scheme, with formation of a cyclometalated dimer intermediate 1, which is cleaved with the addition of the ancillary ligand to form 2. Example ligands are found in Figure 1.11.

result of their strong spin-orbit coupling. This phenomenon permits an electron excited to a singlet state to transfer to a triplet via intersystem crossing. The resulting collection of both singlet and triplet excitons is responsible in part for the high quantuwm yields demonstrated [15]. Relatively high in the spectrochemical series, cyclometalating ligands such as phenylpyridine increase the d − d energy gap, leading to less radiationless quenching, further enhancing the quantum yield. Careful placement of electron donating and withdrawing groups on the cyclometalating ligands tunes the π − π∗ gap, and allows control over the emission wavelength.

N N N S N = N C C C C C

N N N N N = N N N N N N H NH N

Figure 1.11: Various representative cyclometalating and ancillary ligands.

The most common route to an iridium cyclometalate monomer compound 2 is through the intermediate formation and isolation of a dichloro-bridged dimer 1, though recently a one pot synthesis has been demonstrated [41]. The cyclometalation reaction proceeds through the mechanism shown in Figure 1.12. First, a dative interaction forms between a donor

13 atom (here N) and the metal, bringing a C−H bond near the electron-rich metal centre. Intramolecular cleavage of this bond results in the 5-membered metallocycle, and a metal hydride. Subsequent hydride abstraction leaves the metal centre in its initial oxidation state, and the proton bound to a displaced ligand. The process is repeated to yield a bis-cyclometalated compound, which preferentially dimerizes instead of undergoing a third cyclometalation. This mechanism is fully discussed in elsewhere [42, 43, 44]. In the past these reactions have involved ∼20 hour reﬂux steps in glycol ethers such as 2-ethoxyethanol. With the advent of microwave reactors, however, the reaction time has been lowered by an order of magnitude [45, 46]. In addition, water has been shown to adopt a range of solvent properties under microwave radiation, and reactions previously considered non-aqueous have been shown to progress in water [47].

C H C H C H C

+ MLn MLn-1 MLn-1 MLn-2 + LH

N N N N

Figure 1.12: A schematic pathway of the cyclometalation reaction

1.2.3 Radio-Iridium

Despite the popularity of both multifunctional radiopharmacy and iridium cyclometalate compounds as LCI agents, eﬀorts undertaken so far have failed to produce a radio-iridium compound for study. Figure 1.13 shows the iridium isotopes with half-lives of interest decay via electron capture. Further examination reveals the deactivation of the respective daughter nuclides to be convoluted, with many γ rays of little use to imaging. Turning to therapy, 192Ir emerges as a likely candidate with a 95% β− branching ratio, and a 179 keV average emission energy [17]; this potential has been recognized through its use as a high-intensity β− brachytherapy emitter [48]. However, its production in a reactor via neutron activation of an enriched 191Ir wire is incompatible with labelling chemistry due to the low speciﬁc activity obtained. The cyclotron production of 192Ir via the 192Os(p, n) reaction channel has been studied with regard to its utility to the production of brachytherapy sources, but the cross section of the reaction was found to be too low for use in this application [20]. A result of this study, however, was the careful examination of the excitation function, as shown in Figure 1.4, and the yields attainable with thin targets. This data suggests current osmium targetry techniques would provide enough activity to demonstrate a labelling experiment with 192Ir.

14 184Pt! 185Pt! 186Pt! 187Pt! 188Pt! 189Pt! 190Pt! 191Pt! 192Pt! 193Pt! 194Pt! 195Pt! 196Pt! 17.3 m! 70.9 m! 2.08 h! 2.35 h! 10.2 d! 10.87 h! 2.83 d! 50 y! ε: 100% ε: 100% ε: 100% ε: 100% ε: 100% ε: 100% 0.12%! ε: 100% 0.78%! ε: 100% 32.86%! 33.78%! 25.21%!

183Ir! 184Ir! 185Ir! 186Ir! 187Ir! 188Ir! 189Ir! 190Ir! 191Ir! 192Ir! 193Ir! 194Ir! 195Ir! 57 m! 3.09 h! 14.4 h! 16.6 h! 10.5 h! 41.5 h! 13.2 d! 11.78 d! 73.8 d! 19.28 h! 2.5 h! - ε: 100% ε: 100% ε: 100% ε: 100% ε: 100% ε: 100% ε: 100% ε: 100% 37.30%! β : 95% 62.70%! β-: 95% β-: 100% ε: 5% 182Os! 183Os! 184Os! 185Os! 186Os! 187Os! 188Os! 189Os! 190Os! 191Os! 192Os! 193Os! 194Os! 21.84 h! 13.0 h! 93.6 d! 15.4 d! 30.11 h! 6.0 y! ε: 100% ε: 100% 0.02%! ε: 100% 1.59%! 1.96%! 13.24%! 16.15%! 26.26%! β-: 100% 40.78%! β-: 100% β-: 100%

Figure 1.13: A small portion of the table of nuclides, centred on potentially useful iridium isotopes [17]. Isotopes in black are stable, and blue and brown decay by EC () and β−, respectively.

1.3 Thesis Overview

This thesis aims to demonstrate the radiosynthesis of a 192Ir cyclometalate compound for the potential study and application as a multifunctional therapeutic agent. While having demonstrated success in brachytherapy, 192Ir has yet to be applied to targeted radiotherapy. This combined with the success of iridium cyclometalate compounds in LCI has provided a unique opportunity for study. The research was separated into three broad chapters: Target Chemistry, Isotope Production, and Iridium Cyclometalate Chemistry. Explored first is the production of targets suitable for proton bombardment. An electrolyte bath containing OsO4 was used to electroplate thin deposits of natural osmium onto silver target backings. These targets were then irradiated on the TRIUMF TR13 cyclotron, generating a range of iridium isotopes which were characterized and quantified with γ-ray spectroscopy before their separation from the target material. Parallel to this work, a representative cyclometalate chemistry was chosen and developed in the context of a radiochemical synthesis. The culmination of these projects was the first known synthesis and isolation of a radio-iridium cyclometalate compound, whose identity was confirmed with high performance liquid chromatography techniques.

15 Chapter 2

Targetry

The production of 192Ir has been demonstrated via irradiation of an enriched 192Os target [20, 21]. This project seeks to recreate these experiments using natOs as the target material, with the goal of obtaining enough radio-iridium to demonstrate a labelling experiment. To achieve this goal, current procedures will be modified to achieve the thickest targets possible, that are still suitable for bombardment. Robust target design is essential to isotope production by any method, such that the isotopes generated remain in the matrix for the duration of bombardment up until radiochemical workup. A variety of target designs employing all three phases of matter are used to achieve the most efficient generation, retention, and separation of the desired isotope. Given that osmium is a metal, here we investigate the production of solid osmium targets suitable for proton bombardment. A summary of the existing approaches to electroplating osmium on a variety of metallic media is discussed in detail in [49]. Most methods have difficulty achieving deposits thicker than several milligrams per square centimetre without the formation of OsO2. Several authors have explored this topic with regard to robust targets capable of withstanding particle bombardment; successful experiments shared basic conditions, as well as the use of sulphamic acid (NH3SO3) as either an electrolyte enhancement, or as a ligand to OsO4 to form an osmium sulphamate complex [20, 21, 50]. The goal of these experiments is to achieve the thickest target possible, maximizing the potential yield of the iridium isotopes.

2.1 Experimental

2.1.1 Reagents and Instrumentation

Natural abundance osmium powder (99.95%, 20 mesh) and osmium tetroxide (OsO4) were purchased from Alfa Aesar International (MA, USA); 18 M nitric acid (HNO3), sulphamic acid (H3NSO3), sodium phosphate (Na2HPO4), potassium hydroxide (KOH) and sodium

16 hydroxide (NaOH) were purchased from Sigma Aldrich Canada (Oakville, Ontario). Scan- ning electron microscopy and electron dispersion spectroscopy (SEM/EDS) were performed on a FEI Dualbeam 235 in the SFU 4D Labs.

OsO4 is extremely toxic and highly volatile. Careful consultation of the material safety data sheet should precede any work with this compound, which should only take place in a well ventilated fumehood.

2.1.2 Distillation of OsO4

A distillation apparatus was assembled as in Figure 2.1, complete with a distilled water trap. In the distillation ﬂask metallic osmium (500 mg) was suspended on 4 mL of 18 M

HNO3, and in the receiving flask was placed 10 mL deionized (DI) H2O. The distillation flask was brought to a temperature of 180◦C and allowed to reflux to dryness. Meanwhile,

OsO4 vapours were observed to dissolve in the receiving flask, gradually changing the colour of the solution to green. Once dry, an additional 2 mL of 18 M HNO3 were placed in the receiving flask and again allowed to boil dry. KOH was added to the solution to neutralize the extreme acidity resulting from HNO3 distillation, and a pH of 5 measured with indicator −1 paper. The resulting stock solution, with an assumed OsO4 concentration of 26 mg mL , was stored under a ground glass joint and parafilm in a fumehood until required.

2.1.3 Osmium Electroplating

Natural silver was machined into circles 1 mm thick and 35 mm in diameter for use as target backings. Kapton tape was used to mask oﬀ a circular shape 10 mm in diameter in the centre of the target plate, which was placed in the target mask as seen in Figure 2.1.

In a 120 mL beaker containing 30 mL DI H2O, sulphamic acid (1.257 g, 12.9 mmol) was neutralized with KOH (0.601 g, 10.7 mmol), conﬁrmed with indicator paper; then Na2HPO4 (0.205 g, 1.4 mmol) and NaOH (0.207, 5.2 mmol) were added. The ﬁnal addition of 4 mL

OsO4 stock solution changed the colour to deep brown, and when boiled 5 minutes, changed further to murky yellow. Of this solution 10 mL was set aside to replenish the bath while running, and the remainder made up to 60 mL. The target mask was placed in the electrolytic bath and connected to the negative pole of a low voltage power source. A platinum wire was connected to the positive pole as the anode. The voltage and current were slowly increased to achieve values of 3.2 V and 60 mA, respectively, and the circuit was allowed to run for 1.5 hours. The solution was stirred vigorously and the temperature maintained at 75◦C throughout; the replenishing solution was added periodically along with DI H2O to maintain the volume of the bath.

17 The target obtained was rinsed with DI H2O and allowed to dry in air before weighing. A representative sample was imaged and analyzed with SEM/EDS. Using an ad-hoc method to quantify the material adhesion and structure, a piece of Kapton tape was applied firmly to the surface and pulled off, showing minimal osmium removal. Assessment of target thickness was attempted with a profilometer, however the extreme difference in reflectivity between the silver backing and osmium surface rendered this analysis ineffectual.

Baths were also prepared using crystalline OsO4 in several conﬁgurations, attempting to mimic the distillation process, but satisfactory plating was not achieved.

Figure 2.1: A. The distillation apparatus B. Target in target mask in electrolyte bath C. Finished target

2.2 Discussion

Electroplating trials were ﬁrst attempted using natural copper as the backing material, however the zinc isotopes produced as a result of the natCu(p, n) reactions, 65Zn in particular, have several high intensity γ-rays that unnecessarily increase the handling dose [17, 51]. For this reason, natural silver was selected as the target material, having been successfully demonstrated in the production of many radionuclides from solid targets [52]. The modiﬁed method developed here was shown to produce robust targets suitable for proton bombardment, to a maximum mass of 27 mg. The area plated was measured to be 1.09 ± 0.05 cm2, and so a maximum theoretical thickness of 11 µm is calculated. While thicker targets would be preferable for isotope production, extension of various electroplating conditions allowed no extra growth (AppendixA, Table A.1). This may be due to the formation of the structures visible in Figure 2.2, which appear to have achieved a maximum

18 diameter of ∼15 µm. As a side note, these structures may facilitate the radiochemical process by providing a greater surface area for reaction.

These layers are thicker than that achieved in the literature (≤ 3µm), however other authors attempted only thin, shiny deposits. The inability to characterize the thickness of the targets, as well as the non-uniformity observed in the SEM images, introduces uncertainty in the distribution of the plated material, which may lead to inconsistent isotope production yields between targets. While thicker targets are desirable from an isotope production perspective, the targets produced here will suﬃce. If higher isotope yields are required, either beam current or duration may be manipulated to achieve this.

Though alternate methods for target preparation were not explored experimentally, the literature suggests thicker electroplated targets are not possible. Methods involving brazing (pressed powder pellets; electrophoretic deposition) are limited by the extremely high melting point of osmium metal (3300 K). A solution target demands a suitable salt of osmium with high solubility, low chemical reactivity with the target container, and a low proton cross section of the anion.

1.! 2.!

3.! 4.!

Figure 2.2: Scanning electron micrographs of the osmium target material electroplated on a silver backing. Images 1 - 3 show increasing magniﬁcations of the same region. Image 4 shows a region just outside the masked target plating surface.

19 2.2.1 Proton Attenuation in Matter

The attenuation of accelerated, charged particles in matter is a process primarily coulombic in nature. Protons are slowed incrementally as they scatter the electrons of the stopping matter. Repulsive interactions with nuclei are far less probable, and are generally neglected [53]. The energy thus transferred is dissipated mostly as heat in the target material, de- manding a knowledge of the thermal conductivity and melting points of a target composite. It follows that the elemental (or, electronic) composition and number density of the stopping material will influence the number of electronic interactions, and hence, the rate of attenuation. Using the Stopping Range of Ions in Matter (SRIM) and Transport of Ions in Matter (TRIM) software toolkits, the interactions of protons with different target materials were modelled [54]. For these calculations the theoretical density of each material is assumed, as the number density is calculated as a function of the measured mass. TRIM calculations predict full stopping of 12.7 MeV protons in 224 µm of natural osmium; similarly, full stopping is achieved in 381 µm of metallic silver. SRIM calculations with 12.7 MeV protons on a 10 µm layer of metallic osmium predicts a beam attenuation of 0.36 MeV. The bulk of the deposited beam energy is dissipated as heat, and so the heat transfer of the target and backing material must be considered [53]. The relevant heat transfer equations are included in AppendixA, and are applied to two beam scenarios that will be discussed in Chapter3. The first bombardment scenario is 5 µA for 1 hour; the second is 20 µA for 2 hours. The resulting heat transfer raises the temperature of the target plate from ambient temperature to 332.3 K and 435.2 K, respectively. This is well below the melting points of either osmium (3300 K) or silver (1235 K), and so the target is predicted to remain stable for the duration of either bombardment.

2.3 Conclusion

A reliable method was established using basic media and an osmium sulphamate complex to electroplate metallic osmium onto silver target backings. The targets were characterized with SEM/EDS, and the micro-structures observed were shown to consist of osmium with a slight oxygen contamination. Uncertainty in the mass and distribution of the osmium deposits reﬂects the unsuitability of these targets for accurate cross section measurements, further discussed in Chapter3.

20 Chapter 3

Isotope Production

The bombardment of natural osmium and subsequent isolation of radio-iridium is a task that bridges nuclear and chemical studies, requiring careful consideration from both perspectives. Various nuclear reactions of 192Os may be found in the literature [20, 21, 55], but those of 186-190Os remain unexplored; while the isolation of iridium from osmium has been studied [56], a radiochemical separation from a cyclotron target has yet to be performed. The ﬁnal goal of cyclometalation is an important constraint on the separation chemistry as it deﬁnes the required chemical form of the radio-iridium.

Table 3.1: Isotopic composition of naturally abundant osmium

AOs 186 187 188 189 190 192 Abundance (%) 1.59 1.96 13.24 16.15 26.26 40.78

The osmium coatings produced for these experiments were shown in the previous chapter to approximate thin target conditions. SRIM simulations showed 100% transmission of 12.7 MeV protons, and 0.34 MeV energy loss through the osmium layer, leaving the beam to stop fully in the silver backing. Consequently, there are two corresponding groups of nuclear reactions (and products) to consider. The low energy reactions of natural silver, 107,109Ag(p, n)107,109Cd, have well deﬁned cross sections [57, 58]. Of the isotopes found in natural osmium (Table 3.1), however, only reactions of 192Os ((p, xn), x = 3 − 7; (d, 2n)) have been described in the literature [20, 21, 55].

Radionuclidic Separations

Most routes of isotope production involve puriﬁcation of the product, usually from the target material. The chemical properties of bulk materials also apply to tracer concentrations, and

21 n

Cl N

Figure 3.1: Type 2 anion exchange resin consisting of a trimethyl quaternary ammonium group appended onto a styrene divinylbenzene copolymer, in a chloride form. Once equilibrated, the chloride ion will exchange only for an ion having greater aﬃnity for the ammonium group.

many separations use extraction, distillation, or chromatography techniques, among others. Ion exchange materials are typically applied as the stationary phase of a chromatographic column, and are used to isolate a variety of chemical species. Commonly used in the isolation of metal ions are anion exchange materials with trade names Dowex and Bio-Rad. The material used in this work is 1X8, a styrene divinylbenzene copolymer with an appended trimethyl ammonium group (Figure 3.1). Loaded with a chloride counterion, this combination represents one ‘site’ available for exchange with another negatively charged species. In a previous separation of iridium from osmium hydrochloric acid was used as the mobile phase [56]. As the concentration of HCl increases, so does the aﬃnity of the ammonium for the chloride. In this way the column conditions can be tuned to retain or elute a species of interest.

3.1 Experimental

Reagents

All solvents and chemicals were used as purchased from Sigma Aldrich Canada (Oakville, Ont.). Column chromatography was performed with Biorad 1X8 anion exchange material (Hercules, Cal.).

3.1.1 Osmium Target Bombardment

All targets were irradiated on the TRIUMF prototype TR13 cyclotron, as shown in Fig- ure 3.2. The external ion source injects up to 1.5 mA of hydride (H–) ions to the cyclotron chamber, which can then deliver up to 50 µA of 13 MeV protons to two target modules simultaneously. A schematic of the target module is shown in Figure 3.3.

22 Figure 3.2: The prototype TR13 cyclotron at TRIUMF, Vancouver. This self shielded cyclotron is capable of delivering up to 50 µA of 13 MeV protons to 2 target stations simultaneously.

The targets plates indicated in Table A.1 were installed in the target module shown in Figure 3.3. For the duration of the bombardment a stream of helium gas (40 L min−1) was directed at the osmium-coated side, and a stream of water (1 L min−1) at the backside. The ﬁrst experiment consisted of a 10 minute, 5 µA irradiation, followed by a visual in- spection of the target to ensure the stability of the target material, followed by a 60 minute, 5 µA irradiation. No visible degradation of the target was observed after either bombardment, and no activity attributable to isotopes of iridium were found in the cyclotron. For the next two experiments a 5 µA current was applied to the target for 1 hour. The fourth experiment used a 20 µA, 2 hour bombardment, after which contamination was found inside the target holder indicating release of activity from the target. A 24-hour waiting period was observed after each irradiation to allow the decay of short lived isotopes and avoid unnecessary dose.

3.1.2 Gamma-Ray Spectroscopy

All gamma spectra were measured on a shielded, coaxial high purity germanium (HPGe) detector (Ortec, 3.5 keV FWHM at 1174 keV). Three sample geometries were employed to achieve detector dead times <12%: 1) A point source raised 45 cm above the detector; 2) A

23 Coated Target Backing Plate!

Beam! H2O!

Al Beam He! Window!

Figure 3.3: Plan sketch of the target holder employed for all bombardments. Helium gas was directed at the aluminum beam window foil and the front of the target plate at 40 L min−1, and water was directed at the back of the target plate at 1 L min−1.

point source raised 5 cm above the detector; and, 3) A 20 mL vial, raised 1 cm above the detector. A 152Eu point source was used to perform the energy calibration and efficiency correction for geometries 1) and 2), and a standard solution of 152Eu in a standard 20 mL vial was used for geometry 3). The peak-area analysis was done using the Genie2000 software package provided by Canberra, and the activity of each isotope was quantified with the γ-ray emissions from their respective daughter nuclides listed in Table 3.2. After target module disassembly, 24 hours after the end of bombardment, a γ spectrum was collected from each target using geometry 1). Also, the γ-spectrum of the backing plate was collected using geometry 2) after the radiochemical separation to quantify the target material removal efficiency.

3.1.3 Isolation of Radio-Iridium

At each stage of the radiochemical process 5 to 10 µL aliquots were sampled for off-line activity measurements on the HPGe detector. A variety of approaches, including different acids and chemical treatments, were attempted to achieve the dissolution of the target material. For the purpose of brevity, only the successful method is reported here. The irradiated target was heated to 175◦C in air for 10 minutes to oxidize the osmium and iridium. The target material was cleanly dissolved by 100 µL of 35% H2O2, and the resulting solution was collected with a 6 M HCl rinse (2 × 500 µL). This solution was 2– refluxed for 2.5 h, effecting the conversion of osmium and iridium oxides to [OsCl6] and 3– [IrCl6] , respectively. Alternatively, the solution was placed in a microwave reactor for 1.5 h at 120◦C. The resulting solution was then gravity loaded onto a 2 cm (h) × 1.2 cm (d) anion exchange column equilibrated with 6 M HCl. Quickly, 3 × 1 mL aliquots of 6 M HCl were forced through the column and each resulting fraction retained and quantified with a dose

24 7! Ir gamma! Ir x-ray!

6! Cd gamma!

4! Normalized log counts ! Normalized 3!

1! 0! 500! 1000! 1500! 2000! 2500! Energy (keV)!

Figure 3.4: A sample spectrum collected from an irradiated target plate 26 hours after bombardment using an Ortec HPGe detector using a calibrated geometrical conﬁguration. Visible are the characteristic x- and γ-rays from 186-190,192Ir, and the γ-rays from 107,109Cd.

calibrator. The elution chromatogram (Figure 3.5) was constructed in the fourth experiment via collection and measurement of 170 µL eluent fractions on the dose calibrator. The resulting 6 M HCl solutions are incompatible with iridium cyclometalation chemistry, as discussed in Section 4.2. The acidity was neutralized with a saturated solution of

Na2CO3, conﬁrmed with indicator paper.

3.2 Discussion

3.2.1 Irradiation

The TR13 cyclotron is capable of delivering ∼13 MeV protons to the target module. The key variables deﬁning the beam energy seen by the target include: the radius of the stripping foil in the cyclotron, the energy attenuation of the stripping foil, the energy attenuation of the aluminum window foil and, the energy attenuation of the cooling gas. If detailed knowledge of the beam energy and current is required, for example in cross section measurements,

25 1.2!

0.8!

0.6!

! Relative Activity 0.4!

0.2!

0! 0! 1! 2! 3! 4! 5!

Volume (mL)!

3– Figure 3.5: Elution proﬁle of radio-iridium (assumed form [IrCl6] ) from Dowex 1X8 anion exchange material, column dimensions 2 cm (h) × 1.2 cm (d). Activities measured on Capintec dose calibrator.

then monitor foils participating in well described nuclear reactions may be placed in a target configuration [51, 52]. The target configuration used did not accommodate a monitor foil, nor was it essential to the goals of the project to know the precise beam parameters. Therefore SRIM simulations were performed on a 25 µm aluminum foil, and an energy attenuation of ∼0.3 MeV found [54]. The cooling gas was assumed to attenuate the energy negligibly, allowing a beam energy on target of 12.7 MeV. The integrated currents delivered at these energies were 5 and 40 µA·h. The second irradiation condition was necessary to resolve the final target compound, as discussed in Section 4.2.1. Again, without monitor foils, it is difficult to declare the precision of these parameters, and they will be taken at face value. The small contamination discovered in the 40 µA·h run indicated target stability issues which should be addressed before further bombardments at higher currents. For the purposes of this project, the contamination was deemed safe enough to proceed at 20 µA for 2 hours.

3.2.2 Isotope Identiﬁcation and Quantiﬁcation

Irradiation of natural osmium (Table 3.1) with 12.7 MeV protons resulted in the (p, n) trans- formation to the iridium isotopes of corresponding mass number. The cadmium isotopes 107 and 109 were also formed via (p, n) reactions with the natural silver target backing. The

26 Table 3.2: Identiﬁed nuclides generated during proton bombardment of osmium-plated silver. Activity quantiﬁcation was performed using gamma energies, half-lives, and tabulated branching ratios obtained from the Nudat 2.6 database [17].

Nuclide Decay Mode Half-life Daughter Eγ (keV) Iγ (%) 192Ir β−, EC 73.83 d 192Pt, 192Os 316.511 82.86 468.07 47.84 190Ir EC 11.78 d 190Os 361.09 13 371.24 22.8 557.95 30.1 189Ir EC 13.20 d 189Os 245.08 6 188Ir EC 41.50 h 188Os 155.05 30 477.99 14.7 1209.8 6.9 187Ir EC 10.5 h 187Os 177.68 2.2 912.86 4.3 977.54 2.8 186Ir EC 16.64 h 186Os 137.16 41 434.84 33.9 107Cd EC 6.50 h 107Ag 93.12 4.7 796.46 0.065 109Cd EC 461.4 d 109Ag 88.03 3.7

isotopic identities were conﬁrmed by measuring the energies and intensities of the γ-rays emitted by the deactivation of their daughter nuclides, as per Table 3.2. The γ-rays from these isotopes account for all major peaks in the spectra gathered, suggesting negligible, or perhaps short lived, impurities to be present in the target materials. For a complete listing of observed peaks from a sample spectrum, as well as the respective isotopic assignments, see Table A.2 in AppendixA. Decay data for the iridium and cadmium isotopes were taken from the NuDat 2.6 database [17]. The key γ-lines listed in Table 3.2 were referenced by the Genie2000 software to quantify the activity of each isotope. The calculated activities from the two identical 5 µA·h runs, normalized by their respective target masses, are included in Table 3.3. The activity generated by the 40 µA·h irradiation was too high for direct measurement of the target plate, with no geometry available to achieve a dead time < 15%. Following the target dissolution an aliquot was taken and measured, however the nature of the target dissolution process did not allow for quantitative transfer of the dissolved material, and so the yields will not be included here. This data was used instead to establish the radionuclidic purity and eﬃciency of the radiochemical process.

27 To confirm the validity of the yield measurements, the calculated activity of 192Ir, along with target parameters (Table A.1) and irradiation conditions, was used in equation (1.11)) to calculate the cross section of the 192Os(p, n)192Ir reaction for Runs 2 and 3. The values obtained were averaged to yield 46.4 ± 6.2 mb, which agrees well with that published by Hilgers (45 ± 5 mb at 12.5 ± 0.5 MeV [20]) and Szelecsenyi (52.3 ± 5.7 mb at 12.2 ± 0.5 MeV [21]). The uncertainty in this cross section is calculated from several sources, as per equation (1.11): 1) error in calculated activity (2 - 4 %); 2) beam flux variability (5%); 3) error in the foil thickness (7.5%), and 4) detector efficiency (5%). These errors were summed in quadrature for both cross section values, then propagated according to standard techniques to obtain the final error on the average cross section.

Table 3.3: The target mass normalized activities from two 5 µA·h runs, back-calculated to the end of bombardment.

Run 2 Run 3 AIr A (kBq mg−1) ∆AA (kBq mg−1) ∆A 186 29.2 2.3 26.6 1.1 187 323 31 326 23 188 214 16 198 8 189 28.9 2.5 27.1 1.2 190 6.39 0.53 5.82 0.24 192 0.890 0.076 0.973 0.042

As mentioned previously, this experiment was not equipped for proper measurement of cross sections due to the uncertainty in target thickness and distribution (Section 2.2) as well as the uncertainty in beam energy and current (Section 3.2.1). Also complicating this potential measurement are the metastable states of each iridium isotope, described in

Table A.3. With one exception, all of these states are short lived (t1/2 ≤ 3 hours), and any population formed would be decayed beyond detection after the 24 hour cool oﬀ period. 192 The metastable state m2 state of Ir is extremely long lived (t1/2 = 241 years) and has no signiﬁcant γ emissions. In either case, it is not possible to know the populations of these states as they have either fed into the ground state by the time of measurement, or are not measurable anyway. It must be assumed that these metastable states were populated in the irradiation, and potential cross section measurements will have to account for this. For these reasons, the cross section values for the reactions 186-190Os(p, n)186-190Ir will not be reported here. The relatively low yield of 192Ir compared to the isotopes 186 - 190 emphasizes the requirement for enriched target material should this method be adapted to medical use.

28 Despite this, the considerable activity provided by the isotopes other than 192Ir facilitated observation and tracking of the radio-iridium through these experiments.

3.2.3 Radiochemical Separation

Metallic Os, Ir! 1.! o Heat 10 min, ~200 C, cool!

OsO2, IrOx! 2.! Addition of H2O2!

OsO4, IrOx! 3.! Reﬂux in 6 M HCl!

2- 3- [OsCl6] , [IrCl6] ! 4.! 2- Anion exchange: retain [OsCl6] !

3- [IrCl6] ! 5.! Acid neutralized!

3- [IrCl6] ready for cyclometalation!

Figure 3.6: Outline of the radiochemical separation and the corresponding chemical species.

The steps of the radiochemical separation and the corresponding chemical species are shown in Figure 3.6. The low concentration of the iridium species requires that they be 3– described generally, however the species is assumed to be [IrCl6] after the reflux step, as per [56]. Measurement of the target plate γ-spectrum before and after the separation revealed a removal efficiency of 85-95%, however the nature of the activity removal should attach a large degree of uncertainty to this value. The Biorad 1X8 column material has a capacity of 1.2 milliequivalents per millilitre (meq mL−1) of wet resin. A typical column was 2 cm (h) by 1.2 cm (d), a volume of 2.26 mL and a calculated capacity of 2.7 meq. Assuming 100% efficiency of target material 2– removal, 0.027 g of osmium (0.14 mmol) was converted to [OsCl6] , one equivalent of which would displace 1 chloride ion, as per Figure 3.1. Thus, the column as prepared has 2– ∼ 20× the capacity required for retention of the [OsCl6] complex. Assuming the iridium 3– isotopes reacted to completion to form [IrCl6] , measuring the activity of a sample from 3– −6 the pre-column solution leads to the concentration of [IrCl6] in solution, 7.45 × 10 M,

29 −5 3– or 7.45 × 10 mmol of [IrCl6] . This amount is negligible compared to the capacity of the column. Though typically run at a lower acid concentration, the anion exchange column was equilibrated and eluted with 6 M HCl. This was necessary to distinguish between the 2– 3– [OsCl6] and [IrCl6] compounds that would both bind very strongly to the column. With 2– 3– 6 M HCl, the column binds the more positively charged [OsCl6] but not [IrCl6] , as per 2– [56]. Hence, [OsCl6] is retained on the column, visible as a red-orange band at the top, 3– and [IrCl6] is eluted with the profile seen in Figure 3.5. Aliquots were taken from both pre- and post- column solutions and their γ-spectra measured. Comparing the activity measured between the two samples yielded a column efficiency of 87%, qualitatively confirmed by measurement of the activity remaining on the column. This accounts for the second highest loss of activity, after the target dissolution step. The large elution volume may be easily overcome by combining fractions and boiling down; activity measurements before and after boiling suggest no volatile components are present. Through the measurements described here, the overall efficiency of the radiochemical separation was estimated to be ∼80%.

3.3 Summary

A range of iridium isotopes (186-190, 192) were produced from the proton bombardment of natural osmium targets via (p, n) reactions. These isotopes were identified using γ- ray spectroscopy, and their yields quantified. These yields were confirmed through the calculation of the 192Os(p, n)192Ir cross section, and its comparison to published literature values. Beam stopping in the natural silver target plate also resulted in the formation of 107Cd and 109Cd. Measured targets were then oxidized, dissolved, and applied to anion exchange chromatography to isolate the radio-iridium from the osmium target material and target plate contaminations. Colourimetric observation of the column suggested that the osmium was successfully retained, while a column profile measuring the activity eluted clearly showed the presence of radio-iridium at the end of the process. The efficiency of the process was tracked using off-line γ-ray spectroscopy, and the radionuclidic purity of the end product established.

30 Chapter 4

Cyclometalation

Iridium cyclometalate chemistry has been explored extensively with respect to the reactions [14], mechanisms [42], and applications [40]. Notably lacking, however, is the labelling of cyclometalated compounds with radio-iridium, perhaps due to prohibitively long reaction times. The use of microwave reactors has shortened the synthesis time by several orders of magnitude, opening the door to a more practical radiosynthesis [45]. To approach this problem the simplest possible representative compound and synthesis was chosen, as shown in Figures 4.1 and 4.2. Both compounds were first synthesized in macroscopic quantities using stable iridium isotopes, and analyzed extensively for use as High Performance Liquid Chromatography (HPLC) standards. The standard reaction conditions were then modified for compatibility with the cyclotron product iridium and the chemical conditions necessary for its radiochemical process. Radio-analogues were synthesized and their identities confirmed by comparison of the radio-HPLC to the synthesized standards.

4.1 Experimental

4.1.1 Reagents & Instrumentation

Iridium chloride hydrate (IrCl3 · nH2O) and iridium hexachloride (IrCl6) were purchased from Alfa Aesar International (MA, USA); phenypyridine (ppy) and bipyridine (bpy) were purchased from Sigma Aldrich Canada (Oakville, Ontario). All solvents used were HPLC grade unless otherwise stated. Thin layer chromatography (TLC) was performed on aluminum plates coated with silica gel; column chromatography was performed with silica gel (Silicycle, 230 - 400 mesh). All microwave reactions were performed on a Biotage Initiator. A Bruker AV-400 instrument was used to collect 1H NMR spectra at 298 K. Chemical shifts for 1H NMR spectra were referenced to the residual 1H found in the deuterated solvents purchased from Cam- bridge Isotope Laboratories (MA, USA). Spectral processing was performed with Mestrelab

31 Research MestRecNova software. Mass spectra were obtained with the positive ion-mode ESI ion source on an Agilent 6210 time of ﬂight LC/MS instrument. HPLC measurements were performed with an Agilent Inﬁnity 1200 series instrument with a Phenomenex Jupiter C18 column and associated Raytest Gabi NaI detector.

4.1.2 Non-Radioactive Syntheses

Synthesis of Compound 1

IrCl3·nH2O (53.4 mg, 151 µmol) and phenylpyridine (56.7 mg, 365 µmol) were dissolved in a 6:1 mixture of ethylene glycol : de-ionized water, and the solution placed in a microwave reactor for 30 minutes at 135◦C. The bright yellow precipitate formed was isolated by vac- uum ﬁltration, washed twice with water and ethanol, and used with no further puriﬁcation. Yield 64.8% (52.6 mg, 49.0 µmol). 1H NMR (400 MHz, Chloroform-d) δ 9.24 (dd, J = 5.8, 0.8 Hz, 1H), 7.87 (d, J = 7.9 Hz, 1H), 7.73 (td, J = 7.8, 1.6 Hz, 1H), 7.48 (dd, J = 7.8, 1.1 Hz, 1H), 6.79 - 6.71 (m, 2H), 6.56 (td, J = 7.8, 1.4 Hz, 1H), 5.93 (dd, J = 7.8, 0.8 Hz, 1H).

HR-ESI (+)-MS m/z (relative intensity): 501.1 (Ir(ppy)2), 535.1 (Ir(ppy)2Cl).

N N 4Cl- C Cl C H2O Ir Ir 2 IrCl .3H O + 4 + 4H+ 3 2 N 135oC, µwave C Cl C N N 6H2O

[1]

Figure 4.1: Reaction scheme for the synthesis of the dimer compound

(ppy)2Ir(µ−Cl)2Ir(ppy)2.

Synthesis of Compound 2

Compound 1 (44.0 mg, 41 µmol) and bipyridine (14 mg, 9.0 µmol) were dissolved in 2- ethoxyethanol and irradiated in a microwave reactor for 15 minutes at 135◦C. The resulting orange product was puriﬁed by column chromatography. Yield 11.4% (11.2 mg, 16.2 µmol). 1H NMR (400 MHz, Chloroform-d) δ 9.48 (d, J = 8.7 Hz, 1H), 8.27 (t, J = 7.9 Hz, 1H), 7.91 (t, J = 6.7 Hz, 2H), 7.76 (td, J = 7.9, 1.4 Hz, 1H), 7.69 (dd, J = 7.9, 1.0 Hz, 1H), 7.49 (d, J = 5.0 Hz, 1H), 7.42 - 7.37 (m, 1H), 7.04 (td, J = 7.6, 1.1 Hz, 1H), 7.00 (ddd, J = 7.4, 5.8, 1.3 Hz, 1H), 6.92 (td, J = 7.4, 1.3 Hz, 1H), 6.30 (dd, J = 7.5, 0.8 Hz, 1H). HR-ESI (+)-MS m/z (relative intensity): 657.16 (2 - Cl–).

32 X

N N N C Cl C H2O C N Ir Ir + 2 2 Ir + Cl2(g) N 135oC, µwave C Cl C N C N N N N

Cl [2] X PF6 [3]

+ – Figure 4.2: Reaction scheme for the synthesis of [(ppy)2Ir(bpy)] X .

Synthesis of Compound 3

Identical to above, the crude product was synthesized from 1 (44.0 mg, 41.0 µmol) and bipyridine (14.0 mg, 89.6 µmol). Then 2 mL of a 0.85 M solution of NaPF6 (1.70 mmol) was added, and the yellow precipitate formed was washed with 2×5 mL aliquots of H2O and EtOH. The UV-HPLC of the compound conﬁrmed the purity as suﬃcient for use as a UV standard for construction of the calibration curve in Figure 4.4.

Ir + 6 HCl −−−−−−−−→ [IrCl ]3– [4] + 3 H (s) (aq) 110◦C, µwave 6(aq) 2(g)

Figure 4.3: Reaction of iridium black with hydrochloric acid to form iridium hexachloride.

3– Synthesis and Application of [IrCl6]

Iridium black (4.8 mg, 25 µmol) was suspended in 0.5 mL of HCl and reacted in a microwave reactor at 110◦C for 1 hour. Though the iridium did not react to completion, a deep brown solution was formed. The acidity of this solution was neutralized with Na2CO3, conﬁrmed with pH paper, and 20 µL added to DI H2O (180 µL) along with ppy (20 µL, 0.140 µmol). This mixture was reacted in the microwave for 30 minutes at 135◦C. The resulting solution 3– was light brown, indicating incomplete reaction of the [IrCl6] . The reaction solution was analyzed with HPLC under the standard conditions deﬁned in Table 4.1.

4.1.3 Carrier-Added Radio-Synthesis of Compound 2

The following procedure was performed twice, once immediately following completion of the radiochemical separation, and once the following day (28 and 47 hours after end of bombardment, respectively).

33 Table 4.1: High performance liquid chromatography (HPLC) conditions for separation of the cyclometalate compounds of interest. Water and acetonitrile (ACN) were varied with linear gradients at a ﬂow rate of 1.1 mL min−1. Experiments were performed on an Agilent 1200 instrument with a 250 mm Phenomenex Jupiter C18 column.

t (min) % H2O % ACN 0 100 0 2.5 60 40 15 0 100

In a microwave vial, a solution of IrCl3 · nH2O (500 µL, 0.834 mM) was combined with 200 µL of the radio-iridium solution from Section 3.1.3, and an excess of ppy (47.1 mg, 0.303 mmol) added. This mixture was reacted in the microwave at 135◦C for 35 minutes, yielding a yellow organic layer. Bpy (6.1 mg, 0.039 mmol) was added and the mixture reacted in the microwave at 135◦C for 15 minutes, yielding an orange organic layer. This layer was sampled and analyzed with radio-HPLC.

4.2 Discussion

Compound 1 was first synthesized by reacting the cyclometalating ligand phenylpyridine with non-radioactive iridium trichloride hydrate in a microwave reactor (Figure 4.1). Ex- changing the solvent mixture of 3/1 2-ethoxyethanol/water for 6/1 ethylene glycol/water improved the yields found in literature from ∼40% [59] to 65%. After isolating 1, compound 2 was achieved by reaction with the ancillary ligand bipyridine (Figure 4.2). The purity of both compounds was confirmed via 1H Nuclear Magnetic Resonance (NMR) and Mass Spectrometry (MS), and select analyses included in AppendixA. The NMR peak assignments were confirmed via [60]. The samples were then characterized with High Performance Liquid Chromatography (HPLC), and a solvent program (Table 4.1) developed to efficiently separate the two compounds and their respective starting materials. Figure 4.5 shows the clean resolution of compounds 1 and 2, also separate from their starting materials. Despite the recent purchase of the phenylpyridine used for the synthesis of 1, analysis with both NMR and HPLC reveal an impurity, clearly visible in Figure 4.5 at 14 minutes. Difficulty isolating compound 2 in quantities sufficient for further analysis led to the synthesis of 3 in good yield. The low solubility of the PF6 counterion in non-polar and aqueous media easily precipitated the salt, which was shown to have an HPLC time signature identical to 2 when run on the standard program described in Table 4.1.

34 2500!

2000!

1500!

1000! ! Absorption Peak

500!

0! 0.00E+00! 1.00E-05! 2.00E-05! 3.00E-05! 4.00E-05! Concentration (M)!

Figure 4.4: Ultraviolet absorption peak integration - concentration calibration curve for + – [Ir(ppy)2(dpy)] PF6 constructed using HPLC Agilent 1200 with Phenomenex Jupiter C18 column.

Trials also revealed that both reactions progressed using only deionized water as the solvent, and this approach was adopted for the radiosynthesis. At low reaction volumes it was found that ppy and water form a biphasic system; iridium trichloride hydrate from the aqueous layer reacted with ppy at their interface to form 1, which was in turn solubilized by ppy. Similarly, emulsions of water and several oils have been demonstrated to enhance the yields of a variety of molecular imaging agents [61]. Similar to a previous demonstration of a one pot reaction [41], bpy was directly added to this reaction mixture and allowed to react. Compound 2 was observed in the resulting HPLC chromatogram as the dominant product, though 1 was also present along with several unidentiﬁed products. With standard reaction conditions in hand, consideration turned to the dimerization reaction itself. At no carrier added concentration (estimated in using data from Chap- ter 3.2.2 to be ∼0.15 nM), the probability of two iridium atoms ﬁnding each other was assumed to be very low. The concentration of IrCl3·nH2O was scaled down to a level such that the UV-HPLC traces of compounds 1 and 2 were still within the detection limits, and the concentration of the cyclometalating and ancillary ligands were kept constant. In this way a linear calibration curve was developed, bracketing concentrations from 3.60×10−5 to 1.44 × 10−7 M (Figure 4.4), to be used in determining radiochemical yield.

Though most iridium syntheses use IrCl3·nH2O as the starting material, the iridium 3– isotopes are obtained from the radiochemical target process as [IrCl6] in 6 M HCl. To –3 compare the reactivity of IrCl3·nH2O and [IrCl6] , iridium black was reﬂuxed in 6 M HCl

35 10000! ppy! bpy! [1]! 1000! [2]!

100!

10! Absorbance (arbitrary units) ! (arbitrary Absorbance

1! 0! 5! 10! 15! 20! Time (min)!

Figure 4.5: Overlaid UV-HPLC chromatograms of all starting materials and products, obtained using the instrument parameters shown in Table 4.1. Separation is achieved between compounds 1, 2, ppy and bpy, enabling the puriﬁcation of a one pot synthesis. An impurity was observed at 14 minutes in the chromatogram of ppy.

to obtain the hexachloride product as per [56]. This was reacted without purification to obtain a range of products, including one with an identical time signature to 1. Despite poor yield of the dimer compound, estimated to be ∼10% by peak integration, its presence was deemed sufficient to proceed to the radiosynthesis without further procedure modifi- cation. In addition, several trials determined the extreme acidity to be incompatible with cyclometalate chemistry; this was overcome through the addition of a saturated solution of

Na2CO3, which did not appear to eﬀect the yield or purity of the compounds formed.

4.2.1 Radiosynthesis

A Finally, the radio-synthesis proceeded by adding an aliquot of the [ Ir]IrCl6 solution to the aqueous layer of the reaction mixture described above, and the procedure established in the non-radioactive syntheses applied. The results of radio-HPLC of 1 and 2 are shown in Figure 4.6, overlaid on the UV traces of their respective puriﬁed products. A variety of side products are visible in the production of 1, however upon reaction with the ancillary ligand, 2 appears as the dominant product. Increasing the reaction time may increase the A A [ Ir]IrCl6 reacted, however this also sees the formation of [ Ir]Ir(ppy)3 [45]. The speciﬁc activity of 2 is estimated to be 59 ± 9 Ci mmol, found using the calibration curve of the UV reference compound shown in Figure 4.4. The radiochemical purity of the reaction was obtained from the peak integration of the γ-HPLC, ranging from 2 - 68 %. This large

36 variability may be attributed to inconsistent mixing of the organic and aqueous fractions, a hypothesis supported by the varying activity ratios observed between the two fractions across the three trials.

1.6! gamma-dimer1.6! ! gamma-monomer! UV-dimer! UV-monomer!

1.2! 1.2! ! !

0.8! 0.8! Arbitrary Units Arbitrary Arbitrary Units Arbitrary

0.4! 0.4!

0! 0! 0! 5! 10! 15! 20! 0! 5! 10! 15! 20! Time (min)! Time (min)!

Figure 4.6: Overlaid radio and UV chromatograms of compounds 1 and 2, showing the identical time signatures of the radio-iridium labeled compounds and non-radioactive standards synthesized previously.

It should be noted that activity was found and quantiﬁed in both organic and aqueous phases, however, UV-HPLC quantiﬁcation of the organic phase was not possible due to its strong absorption of UV radiation. For this reason, the radiochemical yield was not calculated. To maximize yield, some extraction of the compound from the organic layer may be attempted, in which case it may be simplest to use the original solvent, that is, 2-ethoxyethanol or ethylene glycol.

4.3 Summary

Cyclotron produced radio-iridium was applied to a representative cyclometalation reaction, A + and [ Ir] [(ppy)2Ir(bpy)] identified as the major product. The compound was first synthesized with non-radioactive iridium, and purified as both chloride and hexafluorophosphate salts. Thorough characterization with NMR and MS established the identity of the compound, which was then used to develop an HPLC analysis method for isolation and analysis. The procedure was then modified to achieve compatibility with the radiochemical procedure defined in Chapter3. Additional trials minimized the reaction volume and carrier concentration, necessary to achieve the highest possible specific activity. Finally, radio- iridium was applied to this modified procedure, and the reaction monitored through out with γ-HPLC and offline γ spectroscopy monitoring. With these analyses, the final product was identified and the specific activity and radiochemical purity calculated.

37 Chapter 5

Conclusion

The cyclotron production, radiochemical isolation, and cyclometalation chemistry of radio- iridium is reported in this thesis. The advent of targeted, multifunctional radiopharmaceuticals has generated a demand for varied tools with which to study, image, and treat cancer. The relevance of this work stems from the successes of iridium both as a luminescence cell imaging (LCI) agent, and a therapeutic isotope used in brachytherapy. Thin natural osmium layers were electroplated onto silver backings according to literature precedent and characterized by weighing and SEM. Targets were bombarded with 12.7 MeV protons on the TRIUMF TR13 cyclotron at several current conditions, yielding a range of iridium isotopes that were identified and quantified using γ-spectroscopy. The targets were then dissolved using H2O2 and HCl and purified by anion exchange chromatography. Finally, the iridium isotopes isolated were applied to the synthesis of a representative cyclometalate compound, whose identity was confirmed by comparison with cold-synthesized standards using high performance liquid chromatograpy. Combining the potential of radi- olabeling with the demonstrated flexibility of cyclometalate chemistry opens the door to a variety of applications in molecular imaging and nuclear therapy.

Future Work

The encouraging results of these experiments point to further questions, and room for improvement exists with regard to radiochemical yield and general chemical workup. While the range of iridium isotopes produced in these experiments aided the tracking of activity, future work will have to consider enriched 192Os as a target material to minimize the excessive dose generated by the isotopes 186 - 190. Implicit in this development is a targetry method more suitable for production through thicker osmium deposits and/or higher beam currents. The prohibitive cost of the enriched material would also encourage study into target recycling.

38 With regard to its application, the behaviour of 192Ir cyclometalate compounds may potentially be observed in vivo, as several γ-rays suitable for imaging are produced in the deactivation of the 192Pt daughter isotope. With the ability to image, cyclometalate compounds might then be further explored with more advanced biological targeting, making use of antibodies, peptides, and other vectors.

39 Bibliography

[1] Ralph Weissleder and Mikael Pittet. Imaging in the era of molecular oncology. Nature, 452(7187):580–9, 2008.

[2] Kevin Brindle. New approaches for imaging tumour responses to treatment. Nat. Rev. Canc., 8(2):94–107, 2008.

[3] Simon Ametamey, Michael Honer, and Pius Schubiger. Molecular imaging with PET. Chem. Rev., 108(5):1501–16, 2008.

[4] Ciprian Catana, Daniel Procissi, Yibao Wu, Martin Judenhofer, Jinyi Qi, Bernd Pich- ler, Russell Jacobs, and Simon Cherry. Simultaneous in vivo positron emission tomography and magnetic resonance imaging. PNAS, 105(10):3705–10, 2008.

[5] Joseph Culver, Walter Akers, and Samuel Achilefu. Multimodality molecular imaging with combined optical and SPECT/PET modalities. J. Nucl. Med., 49(2):169–72, 2008.

[6] Frank Rosch and Richard Baum. Generator-based PET radiopharmaceuticals for molecular imaging of tumours: on the way to theranostics. Dalton Trans., 40:6104–11, 2011.

[7] Roger Schibli and Pious Schubiger. Current use and future potential of organometallic radiopharmaceuticals. Eur. J. Nucl. Med., 29(11):1529–42, 2002.

[8] Cathy Cutler, Heather Hennkens, Nebiat Sisay, Sandrine Huclier-Markai, and Silvia Jurisson. Radiometals for combined imaging and therapy. Chem. Rev., 113(2):858–83, 2013.

[9] Vasilis Ntziachristos. Going deeper than microscopy: the optical imaging frontier in biology. Nat. Meth., 7(8):603–14, 2010.

[10] Monya Baker. The whole picture. Nature, 463:977–80, 2010.

[11] Linus Chiang, Michael Jones, Cara Ferreira, and Tim Storr. Multifunctional ligands in medicinal inorganic chemistry - current trends and future directions. Curr. Top. Med. Chem., 12(3):122–44, 2012.

[12] Vanesa Fern´andez-Moreira,Flora Thorp-Greenwood, and Michael Coogan. Application of d6 transition metal complexes in ﬂuorescence cell imaging. Chem. Comm., 46(2):186– 202, 2010.

[13] Craig Montgomery, Benjamin Murray, Elizabeth New, Robert Pal, and David Parker. Cell-Penetrating Metal Complex Optical Probes: Lanthanide Luminescence. Acc. Chem. Res., 42(7):925–37, 2009.

40 [14] Yun Chi and P. Chou. Transition-metal phosphors with cyclometalating ligands: fun- damentals and applications. Chem. Soc. Rev., 39(2):381–88, 2010.

[15] P. Chou and Yun Chi. Phosphorescent dyes for organic light-emitting diodes. Chem. Eur. J., 13:380–95, 2007.

[16] Kenneth Lo and Kenneth Zhang. Ir (III) complexes as therapeutic and bioimaging reagents for cellular applications. RSC Adv., 2(32):12069–83, 2012.

[17] NuDat 2.6 database.

[18] Kris Heyde. Basic Ideas and Concepts in Nuclear Physics. IOP Publishing, Bristol and Philadelphia, 3rd edition, 2004.

[19] Directory of Cyclotrons used for Radionuclide Production in Member States. IAEA, Vienna, 2006.

[20] K. Hilgers, S. Sud´ar,and S. Qaim. Experimental study and nuclear model calculations on the 192Os(p, n)192Ir reaction: comparison of reactor and cyclotron production of the therapeutic radionuclide 192Ir. Appl. Radiat. Isot., 63(1):93–8, 2005.

[21] F. Szelecs´enyi, C. Vermeulen, G. Steyn, Z. Kov´acs,K. Aardaneh, and T. van der Walt. Excitation functions of 186-190,192Ir formed in proton-induced reactions on highly enriched 192Os up to 66 MeV. Nucl. Inst. Meth. Phys. Res. B, 268(20):3306–14, 2010.

[22] L. Szilard and T. Chalmers. Chemical separation of the radioactive element from its bombarded isotope in the Fermi Eﬀect. Nature, 134:462, 1934.

[23] Gopal Saha. Physics and Radiobiology of Nuclear Medicine. Springer Science and Business Media, New York, 4th edition, 2013.

[24] W. Volkert, W. Goeckeler, G. Ehrhardt, and A. Ketring. Therapeutic radionucides: production and decay property considerations. J. Nucl. Med., 32(1):174–85, 1991.

[25] S Carlsson. A glance at the history of nuclear medicine. Acta Onc., 34(8):1095–102, 1995.

[26] Open Source Handbook of Nanoscience and Nanotechnology, 2013.

[27] O. Chiewitz and G. de Hevesy. Radioactive indicators in the study of P metabolism in rats. Nature, 136:754–755, 1935.

[28] J¨urgenWillmann, Nicholas van Bruggen, Ludger Dinkelborg, and Sanjiv Gambhir. Molecular imaging in drug development. Nat. Rev. Drug Disc., 7(7):591–607, 2008.

[29] Shinya Inoue. Foundations of Confocal Scanned Imaging in Light Microscopy. In J Pawley, editor, Handbook of Biological Confocal Microscopy, chapter 1. Springer Science and Business Media, New York, 3rd edition, 2006.

[30] Matthias Nahrendorf, Edmund Keliher, Brett Marinelli, Peter Waterman, Paolo Fer- uglio, Lioubov Fexon, Misha Pivovarov, Filip K. Swirski, Mikael Pittet, Claudio Vine- goni, and Ralph Weissleder. Hybrid PET-optical imaging using targeted probes. PNAS, 107(17):7910–5, 2010.

41 [31] Lucy Jennings and Nicholas Long. ‘Two is better than one’ - probes for dual-modality molecular imaging. Chem. Commun., pages 3511–24, 2009.

[32] Flora Thorp-Greenwood and Michael Coogan. Multimodal radio-(PET/SPECT) and ﬂuorescence imaging agents based on metallo-radioisotopes: current applications and prospects for development of new agents. Dalton Trans., 40:6129–43, 2011.

[33] Karin Stephenson, Sangeeta Banerjee, Travis Besanger, Oyebola Sogbein, Murali Lev- adala, Nicole Mcfarlane, Jennifer Lemon, Douglas Boreham, Kevin Maresca, John Brennan, John Babich, Jon Zubieta, and John Valliant. Bridging the gap between in vitro and in vivo imaging: isostructural Re and Tc complexes for correlating ﬂuores- cence and radioimaging studies. J. Am. Chem. Soc., 126:8598–9, 2004.

[34] Rachel Huang, Graeme Langille, Ravanjir Gill, Cindy Mei, Jin Li, Yuji Mikata, May Wong, Donald Yapp, and Tim Storr. Synthesis, characterization, and biological studies of emissive Re-glutamine conjugates. J Biol Inorg Chem, 18:831–44, 2013.

[35] Galina Loukova, Wolfgang Huhn, Vladimir Vasiliev, and Vyatcheslav Smirnov. Ligand- to-metal charge transfer excited states with unprecedented luminescence yield in ﬂuid solution. J. Phys. Chem. A., 111(20):4117–21, 2007.

[36] Victoria E. Pritchard, Flora Thorp-Greenwood, Rebeca Balasingham, Catrin Williams, Benson Kariuki, James Platts, Andrew Hallett, and Michael Coogan. Simple polyphenyl Zr and Hf metallocene room-temperature lumophores for cell imaging. Organometallics, 32:3566–9, 2014.

[37] Kai Chen, Zi-Bo Li, Hui Wang, Weibo Cai, and Xiaoyuan Chen. Dual-modality optical and positron emission tomography imaging of vascular endothelial growth factor receptor on tumor vasculature using quantum dots. Eur. J. Nucl. Med. Mol. Imag., 35(12):2235–44, 2008.

[38] Kenneth Lo, Wendell Law, Joey Chan, Hua-Wei Liu, and Kenneth Zhang. Photo- physical and cellular uptake properties of novel phosphorescent cyclometalated Ir (III) bipyridine d-fructose complexes. Metallomics, 5:808–12, 2013.

[39] Roger Tsien, Lauren Ernst, and Alan Waggoner. Fluorophores for Confocal Mi- croscopy: Photophysics and Photochemistry. In James Pawley, editor, Handbook of Biological Confocal Microscopy. Springer Science and Business Media, New York, 3rd ed. edition, 2006.

[40] Youngmin You. Phosphorescence bioimaging using cyclometalated Ir (III) complexes. Curr. Op. Chem. Biol., 17(4):699–707, 2013.

[41] S´ebastienLadouceur, Daniel Fortin, and Eli Zysman-Colman. Enhanced luminescent Ir (III) complexes bearing aryltriazole cyclometallated ligands. Inorg. Chem., 50(22):11514–26, 2011.

[42] Alexander Ryabov. Intramolecular activation of C-H bonds in transition-metal complexes. Chem. Rev., 90(2):403–24, 1990.

[43] A. Shilov and G. Shul’pin. Activation of C-H bonds by metal complexes. Chem. Rev., 97(8):2879–2932, 1997.

42 [44] Iwao Omae. Intramolecular ﬁve-membered ring compounds and their applications. Coord. Chem. Rev., 248:995–1023, 2004.

[45] Hideo Konno and Yoshiyuki Sasaki. Selective one-pot synthesis of facial tris-ortho- metalated Ir (III) complexes using microwave irradiation. Chem. Lett., 32(3):5–6, 2003.

[46] Kaori Saito, Noriyuki Matsusue, Hiroshi Kanno, Yuji Hamada, Hisakazu Takahashi, and Takeko Matsumura. Microwave synthesis of Ir (III) complexes: synthesis of highly eﬃcient orange emitters in organic light-emitting devices. Jpn. J. Appl. Phys., 43(5A):2733–4, 2004.

[47] Doris Dallinger and C. Kappe. Microwave-assisted synthesis in water as solvent. Chem. Rev., 107(6):2563–91, 2007.

[48] J. Habrand, A. Gerbaulet, M. Pejovic, G. Contesso, S. Durand, C. Haie, J. Genin, G. Schwaab, F. Flamant, M. Albano, D. Sarrazin, M. Spielmann, and D. Chassagne. Twenty years experience of interstitial Ir brachytherapy in the management of soft tissue sarcomas. Int. J. Rad. Onc. Biol. Phys., 20(3):405–11, 1991.

[49] Terence Jones. Electrodeposition of Os. Met. Finish., 100(6):84–90, 2002.

[50] S. Chakrabarty, B. Tomar, A. Goswami, V. Raman, and S. Manohar. Preparation of thin Os targets by electrodeposition. Nucl. Inst. Meth. Phys. Res. B, 174(1-2):212–14, 2001.

[51] S. Takács,F. Tárkányi, M. Sonck, and A. Hermanne. New cross-sections and inter- comparison of proton monitor reactions on Ti , Ni and Cu. Nucl. Inst. Meth. Phys. Res. B, 188:106–11, 2002.

[52] Cyclotron Produced Radionuclides : Physical Characteristics and and Production Methods. Technical Report 468, IAEA, Vienna, 2009.

[53] Glenn Knoll. Radiation Detection and Measurement. John Wiley & Sons, New York, 4th edition, 2010.

[54] J. Ziegler, J. Biersalc, and U. Littmark. Stopping and Range of Ions in Matter, 2008.

[55] F. Tárkányi, A. Hermanne, S. Takács, K. Hilgers, S. Kovalev, A. Ignatyuk, and S. Qaim. Study of the 192Os(d, 2n) reaction for production of the therapeutic radionuclide 192Ir in no-carrier added form. Appl. Radiat. Isot., 65(11):1215–20, 2007.

[56] E. Campbell and F. Nelson. Rapid ion-exchange techniques for radiochemical separations. J. Inorg. Nucl. Chem., 3:233–42, 1956.

[57] Bruno Linder and Ralph James. Cross sections for nuclear reactions involving nuclear isomers. Phys. Rev., 114(1):322–25, 1959.

[58] J. Wing and J. Huizenga. (p, n) Cross Sections of 51V, 52Cr, 63Cu, 65Cu, 107Ag, 109Ag, 111Cd, 114Cd, and 139La from 5 to 10.5 MeV. Phys. Rev., 128(1):280–90, 1962.

43 [59] Parvej Alam, Inamur Laskar, Cl`audiaCliment, David Casanova, Pere Alemany, Ma- heswararao Karanam, Angshuman Choudhury, and J. Butcher. Microwave-assisted facile and expeditive syntheses of phosphorescent cyclometallated Ir (III) complexes. Polyhedron, 53:286–94, 2013.

[60] Fred Garces and Richard Watts. 1H and 13C NMR assignments with coordination- induced shift calculations of carbon sigma-bonded ortho-metalated Rh (III) and Ir (III) complexes. Mag. Res. Chem., 31:529–36, 1993.

[61] Ryan Simms, Dong Kim, Darren Weaver, Chitra Sundararajan, Megan Blacker, Karin Stephenson, and John Valliant. Emulsion reactors: a new technique for the preparation of molecular imaging probes. Chem. Eur. J., 18(22):6746–9, 2012.

44 Appendix A

Derivations and Characterization Data

Thin Target Production Equation

Write the diﬀerential equation as the diﬀerence between rate of production (Q) and rate of decay of the product nuclide (λN): dN = Q − λN , (A.1) dt dN −λdt = . (A.2) N − Q/λ

Integrate, assuming that when t = 0, the number of product nuclides N0 is zero: Z t Z N dN − λdt = , (A.3) 0 0 N − Q/λ N − Q/λ −λt = ln . (A.4) −Q/λ Isolating the number of nuclides produced: Q N = (1 − e−λt) . (A.5) λ Rearranging Equation (A.5) yields the activity A. Rewriting the rate of production in terms of target nuclei number density n (nuclei cm−2), reaction cross section σ (cm−2), and beam ﬂux φ (particle s−1):

λN = A = nσφ(1 − e−λt) . (A.6)

The numerical density is calculated using the target mass m and area a, and can be modiﬁed for the case of a target material with more than one isotope using the natural abundance mole fraction for that isotope χA: m L n = · , (A.7) a M m L n = χ · , (A.8) A A a M

45 where M is the molecular mass of the target nucleus, and L is Avogadro’s number. Sub- stituting Equation (A.8) into (A.6) requires the isotope activity AA be used in place of A: χ mLσ φ A = A A (1 − e−λt) , (A.9) A aM which may then be rearranged to obtain the cross section of the speciﬁc reaction σA:

aAAM σA = −λt . (A.10) χAmLφ(1 − e )

Heat Transfer Equations

To predict the thermal response of the target to the beam, three regions must be considered as well as the interfaces between them: the water cooled region on one side of the target, the helium cooled region on the other, and the target volume itself. For simplicity, the thermal conductivity values of the target materials will be averaged according to their thickness instead of treating the target as two distinct regions. First, the temperature diﬀerence (∆T ) across the osmium/silver hybrid is described: kA∆T Q = , (A.11) x where Q is the beam power deposited in the target (10 µA × 12.5 MeV = 250 W); k is the thermal conductivity of the hybrid material (400 W m−1 K−1); A is the cross-sectional area of conduction (calculated with the 10 mm beam diameter, 7.85 × 10−5 m2); and x is the thickness of the hybrid material (1 mm). With these values ∆T is 8.7 K. Second, the temperature across the water/osmium boundary is described:

Q = hA∆T, (A.12) with similarly defined variables (Q = 250 W, A = 10 mm2), as well as h, the heat transfer coefficient (assumed to be 25 000 W m−2 K−1), a ∆T of 127.3 K is obtained. The majority of heat will be dissipated by the water cooling, as shown by the respective temperature increases in the next calculation. As such, the target/helium boundary will be neglected. The temperature rise of the water on one side is described as follows: ∆T mc Q = w , (A.13) 60 with m as the mass rate of water (3000 g min−1); c the specific heat of water (4.184 J g−1 −1 K ); and Q defined as above (250 W), the temperature rise of the water ∆Tw is 1.2 K. The temperature rise of the hellium is described by the same equation, substituting a specific heat of 5 J g−1 K−1 and a flow rate of 40 L min−1 (14.4 g min−1), allowing a

46 ∆THe of 54 K. Again, due to the greater rate of heat dissipation provided by the water over helium, this factor will not be included in the ﬁnal temperature calculation. The total temperature change of the target with 5 mA of 13 MeV protons incident is then approximately 8.7 + 127.3 + 1.2 = 137.2 K. When added to ambient temperature (435.2 K), the melting point of both metallic osmium (3300 K) and silver (1235 K) are well above this value.

47 Table A.1: Summary of electroplating conditions.

◦ Trial H3NSO3 KOH NaOH Na2HPO4 OsO4 Vbath T( C) I (mA) V t (min) mOs Notes (mL) (mg) 1 0.211 0.078 0.080 65 30 2 30 Flaky coating, some adherence 2 0.143 0.478 0.044 32 75 35 2.1 35 12 3 0.175 0.525 0.089 39 65 30 2 30 7 Flaky coating, some adherence 4 0.664 0.432 0.136 0.135 0.089 35 75 20 1.8 30 4 Very thin, uniform coating 5 1.000 0.639 0.200 0.205 0.089 35 75 20 1.8 40 6 Non-uniform coating 6 35 73 30 2 90 15 Previous spent solutions combined. Modiﬁed current conditions 7 2.327 1.470 0.420 0.420 0.281 35 75 30 2 50 23 Flaky coating, some adheres well 8 0.908 0.572 0.165 0.167 0.113 65 72 60 3 100 17 Electrolytic bath heated to 90 degrees, held 5 minutes. Smooth, uniform de- posit, no visible oxides 9 0.721 0.453 0.132 0.144 0.091 10 1.895 1.150 0.207 0.220 0.138 200 72 60 3 120 11 lots of oxides

11 2.003 1.250 0.120 0.300 0.443 125 72 65 3.9 240 18 lots of oxides 48 12 1.410 0.904 0.256 0.261 0.172 125 72 60 3 70 6 lots of oxides 13 1.179 0.739 0.212 0.215 0.406 125 75 20 2.5 poor coating 14 0.997 0.630 0.207 0.205 0.133 125 75 50 3 105 22 fairly uniform, imperfections. Expo- sure to water removes half of material. Oxide formation. 15 0.992 0.639 0.220 0.211 0.154 125 77 45 2.3 90 9 Bubble formed due to configuration. Lots of oxides, easily flaking off 16 0.975 0.635 0.230 0.220 0.160 125 75 50 2.1 90 hideous coating

17 1.025 0.640 0.208 0.222 0.133 125 70 50 2 60 4 awful coating. HNO3 added, neutralized w/ KOH simulates ionic content 18 1.019 0.635 0.217 0.199 0.222 125 75 65 2.1 40 3 Rhodium backing. Poor coating. 19 1.006 0.635 0.185 0.196 0.157 125 72 60 2.1 60 6 Uniform in centre, ﬂaky at edges 20 0.800 0.505 0.162 0.161 0.102 60 75 50 2.1 60 14 Uniform coating, gas evolution holes 21 1.608 1.028 0.335 0.325 0.184 60 75 65 2.2 100 27 Replenishing solution employed. 22 0.805 0.515 0.165 0.163 0.097 60 77 40 2.2 90 15 Smooth, uniform coating 23 0.796 0.515 0.165 0.161 0.102 60 77 50 2.2 90 16 Smooth, uniform coating Iridium Isotope γ-Decay Energies

Table A.2: All observed γ ray emissions from an irradiated Os target. Energies and intensities taken from Nudat database [17].

AIr E (keV) ∆E Intensity (%) ∆E

186 137.157 0.008 41 186 296.9 0.03 62.3 1.5 186 420.81 0.03 2.86 0.15 186 434.84 0.03 33.9 0.9 186 441.48 0.11 1.63 0.19 186 705.1 0.4 0.89 0.19 186 712.57 0.1 0.77 0.23 186 729.5 0.4 0.58 0.12 186 933.34 0.04 5.3 0.3 186 1026.7 0.3 1.19 0.04 186 1057.25 0.08 3.06 0.15 186 1187.9 0.4 1.97 0.23 186 1314.4 0.6 2.4 0.6 186 1508.1 0.7 0.93 0.08 186 1647.4 0.6 4.68 0.23 186 1829.2 0.5 0.12 0.05 186 2144.3 0.5 0.33 0.04 186 2399.1 0.5 0.4 0.05

187 177.68 0.07 2.2 0.23 187 314.13 0.08 0.75 0.08 187 384.96 0.08 0.25 0.04 187 400.81 0.09 3.5 0.4 187 485.96 0.07 0.63 0.07 187 491.74 0.07 1.13 0.12 187 501.51 0.07 1.3 0.16 187 522.13 0.08 0.22 0.024 187 576.6 0.07 0.76 0.08 187 651.41 0.05 0.43 0.03 187 654.39 0.06 0.34 0.09 187 715.99 0.05 0.267 0.015 187 747.63 0.07 0.318 0.02 187 799.88 0.05 0.89 0.04 187 841.26 0.09 0.305 0.015 187 977.54 0.08 2.8 0.3 187 1037.81 0.08 0.384 0.019 187 1102.39 0.08 0.169 0.009 187 1112.14 0.08 0.508 0.025

Continued on next page

49 Table A.2 – Continued from previous page

AIr E (keV) ± Intensity (%) ±

188 155.05 0.04 30 3 188 487.7 0.06 0.212 0.024 188 538.06 0.08 0.194 0.023 188 566.59 0.08 0.212 0.022 188 623.75 0.08 0.26 0.03 188 641.59 0.05 0.38 0.05 188 663.4 0.1 0.063 0.01 188 672.51 0.03 1.44 0.1 188 719.55 0.15 0.022 0.007 188 736.56 0.08 0.29 0.03 188 752.09 0.1 0.075 0.012 188 757.21 0.08 0.39 0.04 188 777.93 0.2 0.096 0.018 188 781.9 0.2 0.17 0.03 188 810.6 0.08 0.171 0.019 188 824.34 0.08 1.03 0.1 188 845.03 0.03 0.173 0.018 188 886.2 0.08 0.25 0.03 188 939.57 0.06 0.66 0.06 188 946.98 0.08 0.129 0.014 188 1017.63 0.06 1.06 0.1 188 1096.54 0.06 1.46 0.12 188 1142.52 0.1 0.37 0.07 188 1149.76 0.09 0.53 0.1 188 1174.56 0.04 1.32 0.13 188 1295.44 0.1 0.134 0.013 188 1302.3 0.2 0.33 0.03 188 1322.93 0.11 0.4 0.03 188 1331.95 0.04 0.47 0.04 188 1336.38 0.15 0.14 0.013 188 1349.54 0.15 0.063 0.009 188 1435.48 0.15 1.48 0.12 188 1452.36 0.2 1.06 0.09 188 1457.55 0.15 1.75 0.15 188 1462.7 0.6 0.41 0.04 188 1465.27 0.15 1.35 0.13 188 1530.1 0.15 0.23 0.03 188 1558.66 0.15 0.87 0.07 188 1618.8 0.4 0.49 0.04 188 1652.66 0.1 0.312 0.025 188 1688.08 0.15 0.74 0.06

Continued on next page

50 Table A.2 – Continued from previous page

AIr E (keV) ± Intensity (%) ±

188 1704.9 1.04 0.15 188 1802.04 0.04 0.97 0.07 188 1809.9 0.3 0.34 0.03 188 1957.11 0.22 0.43 0.04 188 2011.47 0.25 0.62 0.06 188 2040.76 0.25 0.49 0.04 188 2099.1 0.4 4.7 0.6 188 2133.7 0.5 0.096 0.012 188 2193.7 0.4 2 0.4 188 2252.09 0.25 0.35 0.03 188 2347.5 0.4 0.65 0.07 188 2460.8 0.2 0.23 0.03 189 245.08 0.09 6 189 275.89 0.1 0.54 0.05

190 186.68 0.04 52 3 190 196.85 0.15 3.4 0.4 190 223.81 0.05 3.74 0.22 190 361.09 0.05 13 0.6 190 380.03 0.12 2.03 0.11 190 431.62 0.07 2.74 0.18 190 447.81 0.08 2.55 0.14 190 726.22 0.08 3.78 0.16

192 295.96 28.71 0.07 192 308.46 29.7 0.07 192 316.51 82.86 192 604.41 8.216 0.019

51 Table A.3: Energy levels, parities, half-lives and decay modes for each of the iridium isotopes and their known metastable states [17].

A π X E (MeV) J Ir t1/2 Decay Mode % 186 0 5+ 16.64 h 100 0 2- 1.9 h /IT 75/25 187 0 3/2+ 10.5 h 100 0.1862 9/2- 30.3 ms IT 100 188 0 1- 41.5 h 100 0.9235 4.2 ms - 189 0 3/2+ 13.2 d 100 0.3722 11/2- 13.3 ms IT 100 2.3332 25/2+ 3.7 ms IT 100 190 0 11.78 d 100 0.0261 1- 1.12 h IT 100 0.3764 11- 3.087 h /IT 91.4/8.6 192 0 73.83 d β−/ 95.24/4.76 0.0567 1- 1.45 m IT 100 0.1681 11- 241 y IT 100

52 Figure A.1: NMR of Compounds 1 and 2. Assignments were conﬁrmed with data from [60].