Studies Towards Purification of Auger Electron Emitter Er

Home , Isotopes of erbium, Preclinical SPECT

Studies towards puriﬁcation of Auger electron emitter 165Er – a possible radiolanthanide for cancer treatment

Master’s thesis

Salla Tapio

Department of Chemistry University of Helsinki 30 November 2018

Tiedekunta – Fakultet – Faculty Koulutusohjelma – Utbildningsprogram – Degree programme

Matemaattis-luonnontieteellinen tiedekunta Kemia Tekijä – Författare – Author

Salla Tapio Työn nimi – Arbetets titel – Title

Studies towards purification of Auger electron emitter 165Er – a possible radiolanthanide for cancer treatment Työn laji – Arbetets art – Level Aika – Datum – Month and year Sivumäärä – Sidoantal – Number of pages

Pro Gradu -työ 11/2018 87 Tiivistelmä – Referat – Abstract

Syöpä on maailmanlaajuisesti yksi yleisimmistä kuolinsyistä ja syöpäkuolemien määrän odotetaan yhä nousevan johtuen väestönkasvusta sekä elämäntavoista, jotka lisäävät syöpäriskiä. Vaikka useita hoitomuotoja on tarjolla, tarvitaan jatkuvasti uusia, entistä tehokkaampia vaihtoehtoja. Kohdennettu radionukliditerapia on sisäisen sädehoidon muoto, jossa sädehoito voidaan kohdistaa selektiivisesti syöpäsoluihin käyttäen radiolääkeaineita. Radiolääkeaineet ovat molekyylejä, jotka sisältävät sekä sädehoitoon sopivan radionuklidin että biomolekyyliosan, joka sitoutua syöpäsolun pinnan proteiineihin.

Uusia tapoja tuottaa ja puhdistaa sisäiseen sädehoitoon sopivia radionuklideja tutkitaan jatkuvasti. Hoidon kohdentamiseen käytettävien biomolekyylien leimaamiseen tarvitaan kemiallisesti ja radionuklidisesti puhtaita radionuklidiprekursoreita. Radionuklideja tuotetaan usein pommittamalla stabiilia kohtiomateriaalia protoneilla, joten kiinnostuksen kohteena olevan radioisotoopin erottaminen kohtiomateriaalista on erityisen tärkeä osa nuklidituotantoa.

Auger elektroneja emittoiva Erbium-165 on potentiaalinen radiolantanidi syöpähoitoon. Sen tuottaminen vaatii Erbiumin kemiallista erottamista joko Holmiumista tai Tuliumista. LN hartsit ovat ekstraktiokromatografiahartseja, jotka on kehitetty erityisesti lantanidierotuksia varten. Tässä työssä tutkittiin niiden sopivuutta Holmium/Erbium ja Erbium/Tulium –erotuksiin. Erotusmahdollisuuksien arvioimiseksi, näiden kolmen lantanidin pidättymiskapasiteetit määritettiin jokaisessa LN hartsissa. Lisäksi suoritettiin pylväserotuskokeita.

Avainsanat – Nyckelord – Keywords

Sisäinen sädehoito, Auger emitteri, radiolantanidi, LN hartsi Säilytyspaikka – Förvaringställe – Where deposited

E-thesis Muita tietoja – Övriga uppgifter – Additional information

Tiedekunta – Fakultet – Faculty Koulutusohjelma – Utbildningsprogram – Degree programme

Faculty of science Chemistry Tekijä – Författare – Author

Salla Tapio Työn nimi – Arbetets titel – Title

Master’s thesis 11/2018 87 Tiivistelmä – Referat – Abstract

Cancer is one of the most common causes of death worldwide and the amount of cancer deaths is expected to increase due to population growth and lifestyle choices that increase cancer risk. Even though there is a great variety of treatment approaches available, new and more efficient options are in high demand. Targeted radionuclide therapy is a form of internal radiation therapy that can be used to selectively target cancer cells. It utilizes radiopharmaceuticals, molecules that contain a radionuclide and a targeting vector part (biomolecule) which can bind to certain entities expressed on cancer cell surfaces.

Ways to produce and purify new radionuclides suitable for cancer therapy are constantly investigated. Radionuclide precursors of high chemical and radionuclidic purity are needed for radiolabelling of the biomolecules used as targeting vectors. Radionuclides are often produced by proton irradiation of stable target material and separation of the radioisotope of interest from the target material is a vital part of nuclide production.

Auger electron emitter Erbium-165 is a potential radiolanthanide for cancer treatment. Its production would require separation of Erbium from either Holmium or Thulium, adjacent lanthanides to Erbium. LN resins are extraction chromatography resins developed especially for lanthanide separations. In this work, the suitability of these resins for Holmium/Erbium and Erbium/Thulium separations was investigated. In order to assess the possibility of separating the lanthanides of interest on LN resins, the retention capacities of the lanthanide were determined on each resin. In addition, column separation experiments were performed.

Avainsanat – Nyckelord – Keywords

TRNT, Auger electron emitter, radiolanthanide, LN resin Säilytyspaikka – Förvaringställe – Where deposited

E-thesis Muita tietoja – Övriga uppgifter – Additional information Acknowledgements

I would ﬁrst like to thank Paul Scherrer Institute and group leader Dr. Nick van der Meulen for giving me the opportunity to carry out my Master’s thesis work in the Radionuclide Development Group at PSI and Dr. van der Meulen for his constant guidance and advice during my thesis work.

I am grateful to MSc Nadezda Gracheva, the co–supervisor of my thesis work, for sharing her knowledge in the ﬁeld of study with me and for her patience in guiding me and answering my many questions.

I would also like to thank other members of the Radionuclide Development Group, Dr. Zeynep Talip and MSc Roger Hasler, as well as all other colleagues and friends at PSI for their assistance and support.

I wish to express my gratitude to my supervisors, Prof. Anu Airaksinen and Dr. Kerttuli Helariutta, and all my friends at the University of Helsinki for their inspiration and support. I would also like to thank Dr. Helariutta for making me believe in myself and encouraging me to pursue a Master’s thesis position abroad.

Finally, I would like to thank my family, my spouse, and dear friends who have always given me strength and courage. I am deeply grateful to them for constantly reminding me what is truly important in life. I would not be where I am today without their endless love and support.

4 Contents

Part I: Introduction 9

1 Available therapies for cancer treatment 9

2 Application of radionuclides in nuclear medicine: historical overview 12

3 Radiopharmaceuticals in nuclear medicine 13

4 Radionuclides for cancer diagnosis and therapy 15 4.1 Radionuclides for diagnosis ...... 16 4.1.1 Imaging modalities for diagnosis ...... 16 4.1.2 Radionuclides for single photon emission computed tomography ...... 19 4.1.3 Radionuclides for positron emission tomography ...... 21 4.2 Radionuclides for therapy ...... 22 4.3 Radionuclides for theranostics ...... 28 4.4 Possible use of Auger electron emitters in targeted radionuclide therapy ...... 29

5 Novel Auger electron emitter for radiotherapy, Erbium-165 32 5.1 Decay characteristics of Erbium-165 ...... 32 5.2 Prospects in production of Erbium-165 ...... 33 5.3 Chromatographic methods for lanthanide separation and their application for Er/Tm, Er/Ho separations ...... 36 5.4 The concept of distribution coeﬃcients ...... 42

5 6 Aim of Thesis 44

Part II: Experimental 45

7 Materials and methods 45 7.1 Retention capacities ...... 45 7.1.1 Sample preparation for determination of retention capacities 45 7.1.2 Sample preparation for ICP–OES measurements ...... 46 7.1.3 Calculations for determination of retention capacities . . . . 47 7.2 Column separation experiments ...... 48

8 Results and discussion 50 8.1 Retention capacities ...... 50 8.2 Column separation experiments ...... 57 8.2.1 Column separation of Ho and Tm ...... 57 8.2.2 Column separation of Er and Tm ...... 61 8.2.3 Er concentration experiment ...... 66

Part III: Conclusions and outlook 67

A Results of inductively coupled plasma optical emission spectroscopy measurements 85

6 List of abbreviations

α–HIBA = α-hydroxy-isobutyric acid

BBE = Biological bystander eﬀect

CT = computed tomography

DNA = Deoxyribonucleic acid

DOTATATE = DOTA-Tyr3-octreotate

DOTMP = 1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraaminomethylene phosphonic acid

DTPA = diethylenetriaminepentaacetic acid

EC = electron capture

EDTMP = ethylenediamine tetra(methylene phosphonic acid)

FDA = United States Food and Drug Administration

FDG = 2-(F-18)ﬂuoro-2-deoxyglucose

HA = hydroxypaptite

HDEHP = bis(2-ethyl-1-hexyl)phosphoric acid

HEH[EHP] = 2-ethyl-1-hexyl(2-ethyl-1-hexyl)phosphonic acid

HMPAO = hexamethylpropyleneamine oxime

H[DTMPP] = bis(2,4,4- trimethyl-1-pentyl)phosphinic acid

HP = High purity

IC = internal conversion

7 ICP–OES = inductively coupled plasma optical emission spectroscopy

LET = Linear energy transfer mCRPC = metastatic castration-resistant prostate cancer

MIGB = iodine-123 meta-iodobenzylguanidine

MRI = magnetic resonance imaging

PET = positron emission tomography

PM = personalized medicine

PRRT = peptride-receptor radionuclide therapy

PSI = Paul Scherrer Institute

PSMA = prostate-speciﬁc membrane antigen

RIT = radioimmunotherapy

ROS = reactive oxygen species

TRNT = targeted radionuclide therapy

SPECT = single photon emission computed tomography

8 Part I: Introduction

1 Available therapies for cancer treatment

Cancer is caused by a gene change (mutation) that makes the abnormal mutated cells to divide and grow faster than normal cells.1 Cancer is currently one of the most common causes of death worldwide and the burden it produces on society is expected to increase in the future. This is due to population growth and aging as well as lifestyle choices that increase cancer risk. In 2012 the number of new cancer cases and cancer deaths worldwide was estimated at 14.1 million and 8.2 million,2 respectively. Globally, lung cancer and breast cancer are the most frequently diagnosed cancer types and leading causes of cancer deaths in male and female population, respectively. A great variety of cancer treatment types are available, however, there is a substantial demand for new and more eﬀective options, especially, for treating very aggressive, fast spreading and highly metastasized forms of carcinomas.2, 3

Modern cancer treatment includes several different approaches that can be essen- tially classified into three different categories: Surgery, radiation therapy, and the use of anticancer drugs (chemotherapy, immunotherapy, hormone therapy, and targeted therapy).3, 4 Usually, different treatment types are also applied simultaneously in order to achieve the most preferable outcome. Surgery is a suitable treatment option for solid tumors that are located in a specific area. Highly metastasized cancers cannot be sufficiently treated by surgery alone. Also, in the case of leukemia and other types of blood cancer, a different class of treatment is obvi-

9 ously required. Chemotherapy is based on the use of drugs that kill cancer cells. It can be used to treat various types of cancer and it is most often applied in combination with surgery or radiotherapy. The eﬀects of chemotherapy are based on the drugs damaging cancer cells during the cell cycle. However, drugs used in chemotherapy are not selective which leads to serious side eﬀects. Chemotherapy damages especially healthy tissues in which cells are constantly renewed, such as skin, bone marrow, and lining of the digestive system.5, 6

In radiotherapy, ionizing radiation is used to kill cancer cells. It can also be used to treat several kinds of cancer and is commonly used in combination with surgery or chemotherapy. In addition, it can be coarsely divided into external and internal radiotherapy. For external beam radiation therapy, highly penetrating radiation is needed. Most often this means electromagnetic radiation (X-rays or γ-photons) but also electron or proton beams can be utilized. The greatest disadvantage of external beam therapy is the inevitable radiation dose absorbed by healthy tissue in addition to the dose targeted to the tumor or other malignancy.3 Moreover, external beam radiation therapy is not suitable for the treatment of metastasized cancer, since the beam is normally focused on a primary tumor and a limited area around it.

When the cancer cells cannot be targeted precisely enough with external beam radiation therapy, it is possible to resort to internal radiation therapy instead. The tumor type determines which type of internal radiation therapy or radionuclide therapy may result in the most beneﬁcial treatment outcome. For a well-deﬁned, solid tumor, it is possible to carry out brachytherapy, a form of internal radiation therapy, in which a sealed radiation source is placed inside the body in a surgical procedure. There can be one or even several radiation sources, usually in the form of small pellets or beads. The sealed source is placed either next to or even inside

10 the area needing treatment.7

The most important requirements for optimal cancer treatment are the high eﬃ- ciency in damaging/killing cancer cells (so-called therapeutic eﬃcacy) and minimal destruction caused in healthy tissue, which can be achieved by resorting to internal radiation therapy and targeted radionuclide therapy (TRNT). The targeting mechanisms in TRNT utilize certain entities expressed on cancer cell surfaces. It is especially suitable for the treatment of radiosensitive tumors such as leukemias and lymphomas. In contrast, solid tumors are not as susceptible to this kind of treatment since they are not as sensitive to radiation. Solid tumors generally also manifest problems when it comes to the internalization of the radiopharmaceutical in the tumor. This is more problematic for greater tumor volumes.8 However, e.g. 90Y has been successfully applied for therapy of solid tumors.9 In peptide- receptor radionuclide therapy (PRRT) receptors on the surfaces of tumor cells are targeted by radiolabeled peptides.10 Radioimmunotherapy (RIT) is based on targeting antigens on the cell surface using radiolabeled antibodies.10

In recent years, a novel concept of personalized medicine (PM) has gained attention in the area of health care. The goal of PM is to provide ”the right drug, with the right dose at the right time to the right patient”, which has a great impact on the eﬃcacy of treatment as well as the outcome.11 In other words, in PM approach, the aim is to be able to predict patient’s response to a certain therapy.12 Nuclear medicine provides excellent means for assessing the possible success of a therapy approach as well as means for targeted therapy.13, 14

11 2 Application of radionuclides in nuclear medicine: historical overview

Already as early as 1911, Hungarian radiochemist George de Hevesy was investigating the possibility of using a radioactive substance as a tracer for the stable substance of the same element. Since reactors and cyclotrons were not yet in- vented, the radionuclides available for research at the time were nuclides from the nature’s radioactive decay chains. All these elements being heavy metals, their biological toxicity restricted the possibilities of testing de Hevesy’s idea of tracers in practice. However, he performed an experiment concerning plant uptake of a radioactive lead isotope (210Pb)15, 16 and eventually also carried out the ﬁrst radiotracer studies in animals with 210Pb and 210Bi.17

An important step towards realizing de Hevesy’s idea of radiotracers was taken in 1934 when Irene Curie and Frederic Joliot succeeded in producing the first artificial radionuclide, 30P, by bombarding aluminum foil with α-particles produced in the radioactive decay of polonium.18 In the process, the scientists also discovered the concept of positron (β+) decay when 30P decayed into stable silicon, even though the existence of β+-particles had been confirmed two years earlier by Carl An- derson and Victor Hess.18, 19 The invention of the cyclotron in the 1930s and the nuclear reactor in the 1940s led to the production of a variety of artificial radionuclides, including ones suitable for the medical application. In 1935 de Hevesy and Danish physician O. Chievitz published the first results of experiments studying the distribution and metabolism of 32P-labelled compounds in rats15, 20 and only one year later 32P was used for treating leukemia.17 This was closely followed by discovery and utilization of 131I, 99mTc, and 89Sr amongst others.17 131I was first used for diagnostic purposes in 1939 and for treating thyroid cancer in 1946.17

12 The first commercial radiopharmaceutical, 131I-labeled human serum albumin, was sold by Abbott Laboratories in the 1950s and Na131I was the first radionuclide- containing compound to receive the United States Food and Drug Administration (FDA) approval for treatment of thyroid conditions, such as thyroid carcinoma and hyperthyroidism.17 Additionally, in the 1950s first steps were taken towards the development of highly sensitive and accurate imaging modalities that are in use today. One of the most important inventions of the decade, however, was undoubt- edly the 99Mo/99mTc generator.17 Consequently, several important 99mTc-labeled compounds were developed over the following decades. In the 1960s and 1970s first radiopharmaceuticals containing 67Ga and 201Tl were investigated. Since the 1980s a wide spectrum of radiopharmaceuticals available both for imaging and therapy has further expanded, making nuclear medicine one of the most prominent areas of healthcare. Currently, the development of 177Lu radiopharmaceuticals is one of the most investigated areas of nuclear medicine.17, 21

3 Radiopharmaceuticals in nuclear medicine

The term ”nuclear medicine” encompasses both diagnostic methods and therapy approaches that utilize internally administered radioactive substances, namely, radiopharmaceuticals. Radiopharmaceuticals are designed in a way that allows them to localize in a particular kind of tissue or cells. This enables targeting a certain type of malignancy and either observe its distribution in the body or treat it with minimum side eﬀects.

Radiopharmaceutical compound consists of three parts: a biomolecule (so-called targeting vector), a linker and a radionuclide (Figure 3.1).22 Radiometals (e.g. 99mTc, 177Lu) need a chelator to link them to the biomolecule. A chelator is part

13 of the biomolecule and most often macrocyclic chelators are used. The biomolecule for a radiopharmaceutical is chosen based on the needed targeting properties. Non- metallic nuclides such as 18F can be covalently bound directly to the biomolecule and without a linker. Some radionuclides can act as radiopharmaceuticals without being attached to a biomolecule. An example is Na131I, used for treating thyroid carcinoma, since iodide naturally accumulates in the thyroid gland.23 Another

223 example is radium chloride, RaCl2. As a calcium-like element, radium, accumulates in bone tissue and is frequently used for bone pain palliation.24

Figure 3.1: The structure of a radiopharmaceutical. The targeting vector is a biomolecule with aﬃnity to the intended target. Non-metallic radionuclides do not need a linker since they can be directly covalently bound to the biomolecule.22

Radiopharmaceuticals are often administered as intravenous, intra-arterial, or in- tracavitary injections to allow them to enter blood stream in order to reach the target.8 Intratumoral administration is also possible. Radiopharmaceuticals are administered in trace amounts meaning they cannot result in significant pharmacological effects in the body.9 There are several quality requirements for radiopharmaceuticals, which are related to e.g. purity, safety, and efficacy of the product. In practice, quality assurance includes several physicochemical and biological tests. Physicochemical tests are applied in order to confirm pH, ionic strength, osmolal- ity, and radiochemical and radionuclidic purity. Biological tests, in turn, confirm that the product is non-toxic, sterile, and pyrogen-free.9 There is a great variety

14 of radiopharmaceuticals that are currently in use in clinics. Some examples are presented in Table 3.1.

Table 3.1: Some radiopharmaceuticals that are currently used in clinics.

Nuclide Radiopharmaceutical Use 18F[18F]-FDG dianostics25 68Ga [68Ga]-DOTA-Tyr3-octreotate (DOTATATE) dianostics26 111In [111In]-pentetreotide diagnostics27 177Lu [177Lu]-DOTATATE therapy28 153Sm [153Sm]-EDTMP pain palliation29 99mTc [99mTc]-DTPA diagnostics30 90Y[90Y]-ibritumomab tiuxetan therapy31

4 Radionuclides for cancer diagnosis and therapy

The radionuclide used in a radiopharmaceutical agent is chosen based on the physical radiation properties such as type and energy of the emitted radiation and nuclide’s physical half-life and on the other hand the in vivo properties of the labeled compound. For diagnostic purposes, nuclides emitting penetrating radiation are suitable, since radiation is to be detected from outside of the body. Instead, for therapeutical purposes, non-penetrating radiation is desirable in order to save as much normal tissue from the eﬀects of radiation as possible and have all of the energy absorbed in areas to be treated.

15 4.1 Radionuclides for diagnosis

4.1.1 Imaging modalities for diagnosis

The aim in clinically applied radionuclide imaging is to obtain information on how a radiopharmaceutical of interest is distributed in the body of a patient, particularly collection thereof in tumors to determine their location and size. Depending on the decay properties (emission characteristics) of the radionuclide incorporated into the administered radiopharmaceutical, imaging can be performed using either single photon emission computed tomography (SPECT) or positron emission tomography (PET).

The principle of SPECT imaging is based on the detection of photons from γ- emitting nuclides with rotating gamma cameras. In SPECT imaging the measurement is performed using 1–3 gamma camera heads. The main components of a gamma camera include a solid scintillation detector, collimator, and an array of photomultiplier tubes (Figure 4.1). The scintillation detector is usually one large scintillation crystal. Several crystal materials are available but NaI(Tl) is most frequently used due to its sufficient detection efficiency, energy discrimi- nation capability, and reasonable costs. Scintillation crystal transforms incident gamma photons into light pulses that are detected by an array of photomultiplier tubes placed behind the scintillation crystal. Photomultiplier tubes convert the light pulses into electric signals that are then amplified and transferred onwards to signal processing electronics. The collimator is a lead plate full of pinholes perpendicular to the crystal plane. It is placed in front of the scintillation crystal to define the incident direction of the detected gamma rays.32

16 Figure 4.1: The principle of SPECT imaging.33 γ-photons are detected in scintillation crystal material that tranforms photons into light pulses registered by an array of photomultiplier tubes. The incident direction of the photons is determined by placing a lead collimator in front of the scintillation crystal. Collimator only lets through photons with direction perpendicular to the collimator plane.

PET requires radionuclides that decay by positron emission (β+ decay). In β+ decay, the positron annihilates when encountering its antiparticle (electron), resulting in two 511 keV γ-photons emitted in opposite directions. PET imaging is based on the collection of these photons on the detectors in coincidence. In PET imaging a ring-shaped array of scintillation detectors is used to collect the annihilation photons (Figure 4.2). On contrary to SPECT imaging, the use of a collimator is not necessary for PET imaging since the coincidence of the two annihilation photons is utilized in determining the origin of the radioactive decay.34 Radionuclides utilized for PET imaging have β+-energies varying over a wide range. However, since the radiation that is being detected is annihilation photons, the β+ energies only aﬀect

17 spatial resolution of the PET measurement. The greater the energy (and range in tissue) of the β+-particle, the further from decay site the annihilation happens, hence, making the spatial resolution poorer.32

Figure 4.2: Principle of PET imaging.35 PET imaging utilizes β+-emitters: after decay β+-particle encounters electron and resulting annihilation photons are detected by ring of detectors. Origin of decay is determined based on the coincidence of these two photons since they are known to emit in opposite directions.

Advantages of PET over SPECT imaging include many times higher sensitivity, higher spatial resolution, and shorter measurement times which make PET imaging eligible for use in faster dynamic studies.32 In contrast, the often longer half- lives of γ-emitting nuclides used for SPECT allows the observation of the in vivo properties of the tracer for hours or even days after the radiopharmaceutical is

18 administered. Another advantage of PET is the possibility for quantitative imaging (measuring tissue activity concentration).36 However, quantitative SPECT scans can be produced with combined modality SPECT/CT cameras.36 Both PET and SPECT scanners are often integrated with magnetic resonance imaging (MRI) or X-ray computed tomography (CT) scanner since MRI and CT can provide anatomic information with high spatial resolution to accompany the functional information aquired with PET or SPECT.32 Ultimately, the use of these imaging modalities is determined by their availability on site. SPECT instruments are still much more common compared to PET instruments due to their signiﬁcantly lower costs.37

4.1.2 Radionuclides for single photon emission computed tomography

One of the most extensively used radionuclides for SPECT imaging is the γ- emitting radiometal 99mTc. Almost 80 % of all radiopharmaceuticals used in nuclear medicine are compounds that contain 99mTc.9 There are several reasons why 99mTc is so favored in diagnostic radiopharmaceuticals for clinical use. First of all, physical properties and radiation characteristics of this nuclide are nearly ideal. It emits photons only at 140 keV, a rather low energy suﬃcient for the detection with gamma-cameras, while applying only a small radiation dose to healthy tissue. Low photon energy also results in better spatial resolution in imaging because scattered photons are eﬃciently absorbed by the collimator of the gamma-camera. The 6 hour physical half-life of 99mTc is suitable both for labeling and in terms of patient radiation dose considerations.9 99Mo/99mTc generators and various 99mTc- labelling kits are commercially available.38 Frequently used SPECT tracers containing 99mTc include e.g. 99mTc-tetrofosmin for functional cardiac imaging39 and

19 99mTc-HMPAO for functional brain imaging.40

After 99mTc, 111In is the next most well-known nuclide used for SPECT imaging. 111In is produced by means of a cyclotron either with the (p,n) nuclear reaction using 111Cd as target material or with the (α,2n) nuclear reaction from 109Ag.41 111In decays by electron capture, emitting mainly two gamma energies useful for imaging, 171 keV and 245 keV. Having a physical half-life of 68 hours, 111In is well suited for use in antibody-based radiopharmaceuticals.22 For instance, 111In-DTPA is used for cisternography.9

67Ga is another radionuclide frequently employed in SPECT. It can be produced with a cyclotron, by bombarding a natZn target with protons.9 67Ga decays by electron capture to stable 67Zn emitting photons on several energies, mainly below 400 keV. Physical half-life of 67Ga is 78.3 h.22

123I is a cyclotron-produced radionuclide which is also used for SPECT imaging. Its physical characteristics are suitable for diagnosis as it decays with a half-life of 13.2 h emitting 159 keV photons without any accompanying β−-radiation. 123I also emits Auger electrons that contribute to the overall dose (mostly in tumors). 123I-MIBG is commonly used for imaging neuroendocrine tumors.42 131I-MIBG can be used for the same purpose and simultaneously for therapeutic purposes43

131 − as well, since I(t1/2 = 8 d) decays by β -emission and also emits γ-photons on several energies.9

Among novel radionuclides for SPECT imaging, 155Tb should be mentioned as it has appropriate nuclear decay properties such as physical half-life of 5.32 d and two γ-energies, suitable for imaging: 87 keV (32 %) and 105 keV (25 %). 155Tb is not yet clinically employed, but several 155Tb-labeled pharmaceuticals have been studied preclinically.44

20 4.1.3 Radionuclides for positron emission tomography

A challenge of PET imaging is that many of the frequently utilized nuclides are rather short-lived. This has to be taken into account when developing e.g. the labeling procedures. However, the commonly used short-lived nuclides such as 18F, 15O, 13N, and 11C can be rather easily incorporated into biomolecules without modifying the properties of the molecule in question.45 The most commonly used tracer for PET imaging is the glucose analogue 2-[18F]ﬂuoro-2-deoxyglucose (FDG).46 [18F]FDG has been used as a tracer for many malignancies such as lung cancer, breast cancer, lymphoma, colorectal cancer and melanoma. Other examples of PET tracers used in clinics include the benzamide derivative, [11C]raclopride which is used for imaging e.g. patients with Parkinson’s disease45 and [13N]NH3 used for imaging myocardial blood ﬂow.47

68 Ga (t1/2 = 68 min) is the most frequently used radiometal in clinics for PET. Many PET radionuclides are cyclotron-produced but 68Ga is obtained from 68Ge/68Ga generator (68Ge is cyclotron-produced though).9 For example, 68Ga-DOTATATE and 68Ga-DOTATOC are frequently employed for the imaging of neuroendocrine tumors.26, 48

A large number of PET nuclides with longer half-lives are available, such as 124I and radiometals 89Zr and 64Cu. 64Cu is a novel nuclide for PET imaging. It has gained interest due to its suitable physical properties: it has a long physical half- life of 12.7 h and it decays by β+ and β−-emission, and electron capture. 64Cu is produced via (p,n) reaction from 64Ni.9 64Cu-radiopharmaceuticals are not in clinics, however, 64Cu-DOTATATE has been proven to be eﬀective in detecting neuroendocrine tumors in clinical studies.49–51

152 Tb (t1/2 = 17.5 h) has also been proven to be a promising candidate for PET

21 imaging.52 152Tb-DOTANOC has been preclinically studied52 and ﬁrst-in-human study has already been carried out with 152Tb-DOTATOC.53

4.2 Radionuclides for therapy

There are several requirements for radionuclides utilized in TRNT. Most important is that decay characteristics of the radionuclide have to be such that radiation will signiﬁcantly damage cancer cells.8, 10 Another property to be considered is the physical half-life of the nuclide. The half-life should be comparable to in vivo pharmacokinetics of the carrier molecule10 and long enough to ensure that there is time for preparation, transportation, and injection of the radiopharmaceutical. In addition, there has to be enough time for the nuclide to be delivered into the desired biological target. From this point of view, optimal nuclide half-life is between a few hours and a few days. There are some exceptions but having a nuclide with a very short half-life leads to additional requirements like having an on-site

89 cyclotron. Using a nuclide with a long half-life, such as Sr (t1/2 = 50 d) adds to the overall absorbed dose of the non-targeted tissues and may also pose additional challenges regarding radiation safety.8, 10 In addition, to the nuclide having suitable physical characteristics, nuclide production route should be eﬃcient and aﬀordable for the nuclide to be a reasonable choice for therapeutical application (as well as application in diagnosis). Some important radionuclides used for therapeutical applications include 131I, 32P, 89Sr, 90Y, 177Lu, and 225Ac.

Linear energy transfer (LET) value is essential when assessing potential biological damage caused by radiation since it determines the density of ionizations produced by radiation. LET value for α-particles is ca. 80 keV/µm and signiﬁcantly lower for β−-particles in the range of 0.2–2.0 keV/µm.8 The Auger electrons provide low (≤ 100 keV) energies and high LET values in the range of 4–26 keV/µm.8 For

22 β−-particles it is important to optimize the energy according to the tumor size to result in an eﬀective therapy. Energies over 1 MeV are suitable for treating solid tumors since the range of high-energy β−-particles in biotissue can exceed 10 mm.8 In contrast, for small metastases as well as for treatment of leukemias and lymphomas energies less than 1 MeV are adequate.8 Choosing the energy in the case of α-particles or Auger electrons is not similarly essential since their range is always short due to the high LET values. Range of α-particles in biotissue is ca. 50–100 µm.8 LET values determine to what extent ionizing radiation can penetrate tissue and cause DNA damage (and hence lead to the death of a cancer cell) (Figure 4.3). High LET radiation is more likely to cause DNA double strand breaks due to the high ionization density. Low LET radiation, being less densely ionizing, it does not cause DNA double strand breaks so easily. Any ionizing radiation can, however, also cause DNA damage indirectly (Figure 4.3). This happens due to the action of free radicals that are produced in the radiolysis of cellular water. These free radicals, such as OH* and H*, and O2 and H2O2 are often referred to as reactive oxygen species (ROS) (Figure 4.3).8, 10

23 Figure 4.3: All types of ionizing radiation can cause DNA damage via both direct and indirect mechanisms. Direct damage happens when ionizing radiation directly causes chemical bonds within DNA to break. In the case of indirect damage, ionizing radiation produces reactive oxygen species (ROS), radicals that are formed in radiolysis of cellular water. These radicals can react breaking chemical bonds in DNA. Figure from Gudkov et al.8

Another kind of mechanism for radiation-induced cell death is biological bystander eﬀect (BBE). It is a process, where the irradiated cells pass on manifestations of damage to other cells that have not been targeted by radiation. This can result in cell damage and death among the non-irradiated cells. All types of ionizing radiation may induce BBE.54–56

Currently, most of the clinically-employed therapeutic nuclides are β−-emitters, since they are available at various energies and half-lives. However, also α-emitters are being increasingly utilized. In recent years, α-emitter 223Ra has been exten-

223 R sively studied and is currently employed in clinics as RaCl2 (Xoﬁgo ) for the treatment of the bone metastases from the metastatic castration-resistant prostate cancer (mCRPC). Since 223Ra acts as a calcium-analog, the treatment targets bone

24 tissue, especially areas where new cells are actively formed, like bone metastases. The average α-energy of 223Ra is 5.7 MeV and taking into account all its α/β−- emitting daughter nuclides the sum of the decay energies of the complete decay chain is 28.2 MeV. This combined energy is more than 100-fold compared to the average decay energy of another, β−-emitter containing bone-targeting agent, 153Sm- EDTMP. 223Ra is readily available from 227Ac/227Th-generator systems (227Th decays into 223Ra). Also, the rather long physical half-life of 223Ra (11.4 d) allows for long-distance transportation.24, 57–59

Another α-emitting radionuclide, 225Ac, has gained attention recently. The useful- ness of 225Ac-labeled compound against several types of cancers has been demonstrated in preclinical as well as phase I studies.60–63 The first-in-human treatment of mCRPC with 225Ac-PSMA-617 was reported by Kratochwil et al.64 in 2016. 225Ac was proven to be very effective in the treatment of mCRPC, however, side effects remain in question. 225Ac can be produced by separating it from 229Th source or by the nuclear reaction from 232Th or 226Ra. It has several α/β−- emitting daughter nuclides that contribute to the total radiation dose caused by 225Ac.65

Antitumor effects of 213Bi have been confirmed in preclinical and clinical studies.60, 66, 67 DOTATOC labeled with 213Bi has been proven effective against neuroendocrine tumors66 and 213Bi-lintuzumab has been successful in treating patients with acute myeloid leukemia.67 213Bi decays by emitting both β− (98 %) and α- radiation (2 %). It has a short physical half-life (45.59 min) but can be obtained from 225Ac/213Bi generator.60

212Pb-labeled antibodies have been preclinically and clinically investigated. 212Pb

68 − 212 68 212 (t1/2 = 10.64 h) decays by β -emission to Bi (t1/2 = 60.55 min). Bi is an α-emitter that considerably adds to the radiation dose from 212Pb and hence, can

25 also enhance the overall therapeutic eﬀect. 212Pb is obtained from the 224Ra/212Pb generator, but the availability of 224Ra is dependent on the availability of 228Th, which is extracted from spent nuclear fuel.60

131 68 − I(t1/2 = 8.0252 d) decays by β –emission that is accompanied by γ-emission at three energies (364 keV being the most intense). Hence, 131I is suitable for use in both therapy and imaging. It is available as Na131I and is mainly used in diagnosis and treatment of various thyroid conditions, as previously mentioned.9

32P is a pure β−–emitter with a physical half-life of 14.26 d. Its mean β−-energy is 695 keV.68 32P can be produced both in cyclotrons and in reactors. The most common reaction used for production is 32S(n,p)32P. After dissolving and further processing of the target nuclide is obtained in the form of sodium orthophosphate which is used as such e.g. for the treatment of leukemia.9

89Sr is a β−-emitter with a long half-life of 50.6 d and 0.58 MeV mean β−-energy. It is produced with (n,γ) reaction from 89Y. Having chemical properties similar to calcium, strontium accumulates to bone tissue. Due to this property and its physical characteristics, it has been mostly applied for palliation of bone pain caused by metastases.9 Nowadays, the selection of nuclides suitable for this purpose is increasing and 89Sr is being largely replaced by e.g. 153Sm, 223Ra, and 177Lu.24, 69–72

90 − 90 90 Y decays by β -emission (t1/2 = 64.1 h) and is obtained from Sr/ Y generator. β−-particles emitted by 90Y have an average energy of 0.935 MeV which corresponds to a tissue range of 2.5 mm.73 90Y can be incorporated into various kinds of molecules and is hence used to treat numerous types of cancer.9

Currently, the ”gold standard” for radionuclide therapy is clinically-employed 177Lu

− 177 (t1/2 = 6.7 d), which decays by β –particle emission. Lu has also γ-emission

26 on several energies (e.g. 208 keV and 113 keV), which are suitable for SPECT imaging. This allows evaluation of targeting, pharmacokinetics, and excretion of radiopharmaceuticals labeled with 177Lu. 177Lu is produced at high speciﬁc activities in nuclear reactors.28 Attempts to use 177Lu on therapeutical purposes have been reported already since 1960. Examples of potential 177Lu-labeled radiopharmaceuticals include177Lu-DOTATATE (Lutathera R ) for the treatment of neuroendocrine tumors, 177Lu-CC49 monoclonal antibodies for the treatment of adenocarcinoma, 177Lu-J591 monoclonal antibodies for prostate cancer, 177Lu-Anti CD-20 monoclonal antibody for non-Hodgkin’s lymphoma, and the most recent addition, 177Lu-PSMA for metastatic prostate cancer.74, 75

The possibility of using other radiolanthanides besides 177Lu for TRNT is gaining more and more interest. This is due to their diverse decay characteristics as well as the chemical similarities between diﬀerent lanthanides which may allow the use of identical targeting vectors and labelling procedures.76 Lanthanides are well suited for biological applications due to the high redox stability and suitable coordination chemistry of Ln3+ ions.74

Over the recent years, several isotopes of terbium have drawn attention as potential nuclides for therapeutical applications. The suitability of 149Tb for α- therapy as well as theranostic applications has been assessed and results so far show promise.76–78 161Tb has been considered as a therapeutic lanthanide and a potential alternative to 177Lu. 161Tb emits 16 times more Auger electrons per decay than 177Lu which may result in higher therapeutic eﬃciency.79 Indeed, Müller et al. have proved that 161Tb shows higher therapeutic eﬃciency, compared to 177Lu.80

There is a great variety of other β−–emitting radiolanthanides as well, such as 169Er, 153Sm, 143Pr, 149Pm, 165Dy, and 166Ho. In addition, several Auger elec-

27 tron emitting radiolanthanides for therapeutic applications have been investigated. These include e.g. 153Sm, 157Eu, 161Ho, 165Er, and 167Tm. 167Tm can also be utilized for imaging purposes.76 However, the eﬀect of Auger electrons (without other accompanying radiation) on tissues needs to be studied by using pure Auger electron emitters. The possible use of Auger electron emitters in TRNT will be further discussed in section 4.4.

4.3 Radionuclides for theranostics

Theranostics is a method which utilizes two different radionuclides of the same element with the same targeting vector for both imaging and therapy purposes (”matched pair” concept). In some cases it is also possible to use the same nuclide for both imaging and therapy. However, there are only a few such nuclides available. For both of these approaches, the chemical properties of imaging and therapy agent are identical which leads to the reliable predictions of the biodistribution and dosimetry for therapy.81 Due to the lack of better options, nuclides of different elements are still most often used for imaging and therapy. It is important that the chemical properties (especially, coordination chemistry) of these nuclides are similar to each other. 68Ga, for instance, can be applied for imaging, when therapy is to be carried out with 90Y or 177Lu.82 When the chemistry of the two nuclides is different, making predictions regarding the benefits or dosimetry of therapy based on the results of imaging may be unreliable and even hazardous.81, 83–85 Hence, there is a great demand for suitable matched pairs for theranostics. Several elements have been considered for application in theranostics with copper, scandium, and terbium being among the most promising options.86

67Cu is suitable for theranostics since it emits low energy β−-radiation as well as γ photons at 184 keV, 93 keV, and 91 keV. In addition, its physical half-life of 2.6

28 d and versatile labeling chemistry make it an interesting nuclide for theranostics. 67Cu can be paired with 64Cu which is a β+-emitter. Hence, this would allow also PET imaging.81

47Sc emits both low energy β−–particles (average energy 162 keV) and γ-radiation (159 keV) and can be utilized in theranostics. Analogously to67Cu, 47Sc has a suitable pair nuclide for PET imaging, 44Sc. The half-lives of 47Sc and 44Sc are 3.4 d and 4.1 h, respectively. The half-lives would allow transportation of Sc-labeled radiopharmaceuticals to hospitals that do not have a cyclotron to facilitate on-site production of scandium isotopes.87

Another very attractive element for application in theranostics is terbium. Ter- bium has several interesting isotopes. As previously mentioned, 155Tb and 152Tb are suitable for SPECT and PET imaging,44, 53 respectively, whereas α-emitter

149 − 161 Tb (t1/2 = 4.1 h, Eα = 3.97 MeV) and β /Auger electron emitter Tb (t1/2 = 78, 88, 89 6.9 d, Eβ,mean = 0.15 MeV) are possible radionuclides for therapy.

4.4 Possible use of Auger electron emitters in targeted radionuclide therapy

Auger electrons were first reported in 1925.90 These low energy electrons are emitted by atoms that have a vacancy on their inner electron shell. Several processes can result in such a vacancy, typical examples being radioactive decay by electron capture (EC) or internal conversion (IC). This primary vacancy leads to a chain reaction of both radiative and non-radiative transitions where electron vacancies are filled and created, finally resulting in a cascade of emitted low energy electrons (Figure 4.4). Auger electron emitting nuclides can emit on average even tens of electrons per decay.

29 Figure 4.4: Electron vacancy created as a result of electron capture. Vacancy is ﬁlled by an electron from upper shell and excess energy is emitted from the atom in the form of X-ray photon or Auger electron. Figure adapted from91

In general, the energies of Auger electrons vary between only a few eV and ca. 100 keV.92 These energies correspond to a range of less than one nanometer to hundreds of micrometers in water.92 The energies of Auger electrons are very low compared to other types of particle radiation. However, they deposit their energy in a small area which can result in signiﬁcant ionizing eﬀects. In addition to the mechanisms covered in section 4.2, biological damage caused by Auger electron emitters may additionally occur due to the neutralization of the positive charge formed as a result of the decay and emitted electron cascade. The main challenge in utilizing Auger electron emitters in TRNT is how to target malignancies accurately enough to overcome the problem presented by the short range of the Auger electrons.92

Ionizing effects of Auger electrons were first verified in 1963 in a study concerning effects of the decay of 125I, covalently bonded to small hydrocarbons.93 As a result of 125I decay there would be significant fragmentation and ionization of methyl and

30 ethyl groups bonded with 125I. These results aspired interest in studying the possible radiotoxicity of 125I and the effects of this nuclide, especially, when covalently bound to nuclear DNA. 125I along with 123I, 77Br, and 80mBr was proven to be highly radiotoxic expressing effects typical for high-LET radiation.94–97 In subse- quent studies it was also concluded that 125I and other Auger electron emitters are highly radiotoxic even when not covalently bound to nuclear DNA but delivered to close proximity of nuclear DNA.92 Successful delivery of the radionuclide in the cell nucleus is one of the main challenges of TRNT with Auger electron emitters. However, it has been shown for 125I that due to BBE, not every tumor cell needs to be targeted in order to cause a significant decrease in tumor cell growth among non-irradiated cells.98–100 Several options for delivering Auger emitter into the cell nucleus have been considered, one of the most important options being the utilization of thymidine analogs that contain the radionuclide and then bind covalently to the cell nuclear DNA.92

Auger emitters possible for therapeutical applications include e.g. 125I, 123I, 111In, 67Ga, 99mTc, and 201Tl.101–103 However, most of them (123I, 111In, 67Ga, 99mTc, and 201Tl) also emit intense γ-radiation and the physical half-life of 125I is much too long at 59.4 d. Hence, other Auger electron emitting nuclides with more optimal decay properties and especially pure Auger electron emitters are needed for therapeutical applications.

31 5 Novel Auger electron emitter for radiotherapy, Erbium-165

Radioactive isotopes of erbium have raised interest for decades but only few compounds labeled with diﬀerent erbium isotopes have been studied. The suitability of 165Er-citrate for bone imaging and tumor localization was studied already in 1970s.104 Enteric coated tablets containing 171Er have been used to investigate their transit through gastro intestinal tract in humans. Erbium was incorporated on the tablets as stable 170Er and then activated by (n,γ) reaction.105 The possibility of using 169Er labeled compound for radiation synovectomy and bone pain palliation has also been investigated. For this purpose, Chakravarty et al. labeled hydroxypaptite (HA) particles and 1,4,7,10-tetraazacyclododecane-1,4,7,10- tetraaminomethylene phosphonic acid (DOTMP) with 169Er.106

5.1 Decay characteristics of Erbium-165

165Er is an appealing candidate for Auger electron therapy due to several beneficial properties. The physical half-life of 165Er (10.36 h)68 is suitable for medical applications, giving time for the processing of the radionuclide and preparation of the radiopharmaceutical. 165Er decays by electron capture to the stable 165Ho, meaning the decay product does not add to the radiation dose inflicted by 165Er. 165Er emits Auger electrons (5.3 keV, 65.6 %; 38.4 keV, 4.8 %),68 without accompanying gamma radiation. Emission of the low energy X-rays (47 keV – 55 keV)68 allows preclinical SPECT imaging. Based on the above mentioned properties, 165Er could be an appropriate candidate for therapeutic application. However, the effect of Auger electrons on the tumor cells first needs to be investigated in a preclinical

32 setting.

5.2 Prospects in production of Erbium-165

There are several possible production routes for 165Er utilizing both reactors and cyclotrons. In reactors, 165Er can be produced via neutron activation from stable 164Er. However, natural abundance of 164Er is very low and enriched materials are available with only 51–75 % of 164Er. This option is not suitable for production of 165Er for medical applications since the product will contain natural erbium isotopes (e.g. 166Er, 167Er, 168Er, and 170Er) which will result in non-isotopically pure carrier-added product and low radiolabeling yield.104 To obtain carrier-free 165Er cyclotron production routes could be utilized. Possible charged particle production routes of 165Er include direct routes from natural Ho targets as well as indirect routes from Er targets (Table 5.1). Nuclear reactions towards the production thereof include 165Ho(p,n)165Er and 165Ho(d,2n)165Er as well as indirect routes natEr(p,xn)165Tm → 165Er, natEr(d,xn)165Tm → 165Er, and 166Er(p,2n)165Tm → 165Er.107–111

Table 5.1: Possible Erbium-165 production routes, experimentally determined maximum reaction cross-sections and corresponding beam energies.

Production route Target Beam Max. measured material energy (MeV) cross-section (mb) 165Ho(p,n)165Er 165Ho 9.5 ± 0.6 172 ± 19107 165Ho(d,2n)165Er 165Ho 13.6 ± 0.7 597 ± 67108 natEr(p,xn)165Tm → 165Er natEr 22.9 ± 0.4 456 ± 52109 natEr(d,xn)165Tm → 165Er natEr 27.7 ± 1.0 628 ± 71110 166Er(p,2n)165Tm → 165Er 166Er 15.1 ± 0.2 894 ± 104111

33 165Ho is the only natural holmium isotope, which implies that holmium targets for direct production routes are of high availability and low cost. The advantage of these production routes is the possibility of using medical cyclotrons for 165Er production due to the required beam energies.107, 112 However, due to reported low production cross-sections,107 the production yield could be low. According to experimental studies by Tárkányi et al.,107, 108 maximum cross-section values ob- tainable for 165Ho(p,n) and 165Ho(d,2n) reactions are 172 mb and 597 mb, with proton and deuteron beam energies of 9.5 MeV and 13.6 MeV, respectively. Her- manne et al.112 also obtained experimentally a cross-section value of ca. 590 mb for the 165Ho(d,2n) reaction with deuteron beam energy of 12.4 MeV. The value obtained by Tárkányi et al. to the 165Ho(p,n) reaction is supported by computational results by Sadeghi et al.113 However, in the same study a maximum cross-section value of only ca. 300 mb was determined for the 165Ho(d,2n) reaction.

A possible starting material for 165Er production via indirect routes is natural erbium. Since natural erbium consists of six stable isotopes,114 proton bombardment of a natural erbium target produces several other isotopes of thulium in addition to the nuclide of interest (165Tm). These other nuclides, such as 163Tm, 166Tm, 167Tm, and 168Tm then decay by electron capture to various erbium isotopes. As a result, obtaining carrier-free 165Er using this method is not possible. Accord- ing to Sadeghi et al.113 the maximum cross-section value for natEr(p,xn)165Tm reaction is ca. 550 mb with 22 MeV proton beam energy, theoretical yield being 210.9 MBq/µAh. However, Tárkányi et al.109 obtained experimentally a maximum cross-section value of ca. 460 mb for the same reaction with proton beam energy of 23 MeV.

The use of the enriched 166Er gives the advantage of producing carrier-free 165Er via the indirect route 166Er(p,2n)165Tm → 165Er. In two separate computational

34 studies the maximum cross-section for 165Tm production via this reaction was estimated to be at 21 MeV proton beam energy. Sadeghi et al.113 predicted the maximum value for cross-section to be 1263.3 mb, while Zandi et al. determined a value of 1234 mb.115 In these two studies, the best incident proton energy range for 165Tm production was concluded to be 16–23 MeV113 and 12-27 MeV,115 respectively. Tárkányi et al.111 obtained experimentally a maximum cross-section value of 894 mb with 15.1 MeV proton beam energy. For the maximum reaction yield, however, energies ≥30 MeV are necessary which requires the use of larger- scale cyclotron facilities (e.g. PSI’s Injector 2 cyclotron). While medical cyclotrons up to 30 MeV can be utilized for this type of irradiation, the use of larger-scale cyclotron facilities will result in higher production yield.109

The cyclotron facility at Paul Scherrer Institut contains three consecutive accel- erators, accelerating through each to obtain a final proton beam energy of 590 MeV. The first accelerator, a Cockcroft-Walton column, is a type of linear accelerator that produces a proton beam pre-accelerated to an energy of 870 keV. After pre-acceleration the proton beam is transported to Injector 2, a separated-sector cyclotron, which gives additional acceleration resulting in beam energy of 72 MeV (at 2.5 mA). The final acceleration step is provided by an 8-sector ring cyclotron that increases the proton beam energy to 590 MeV.116 The 72 MeV proton beam from Injector 2 feeds the ”Isotope Production” beam line and target station at a beam intensity up to 50 µA, which is sufficient to enable production of 165Er via the chosen indirect production route.

35 5.3 Chromatographic methods for lanthanide separation and their application for Er/Tm, Er/Ho separations

There is an increasing number of radiolanthanides which are used in clinics (177Lu, 153Sm etc.). Radiolanthanides are produced by neutron or proton irradiation of the required target material. Hence, for further preclinical use of the produced radiolanthanide, it is necessary to separate the product from its target material and possible co-produced impurities, as well as preferably extract it in a small volume of diluted acid. The ﬁnal product should be of high chemical and radionuclidic purity, therefore, puriﬁcation is a vital part of the radionuclide processing.9

Separation of the neighboring lanthanides is one of the most challenging areas of ion exchange chromatography117 due to their mutually similar physical and chemical properties, such as similar ionic radii and coordination chemistry.118

A great number of the lanthanide separation procedures have been developed over decades as stable isotopes of lanthanides found their application in various ﬁelds such as metallurgy, petroleum, textiles and agriculture. Until the 1960s, lanthanide separation methods mostly relied on ion-exchange. Later, ion exchange had been partly replaced by solvent extraction processes in applications that require handling of very large volumes of solutions.119, 120 Ion exchange, however, is the method of choice for chemical separation procedures, especially for producing small quantities of high purity lanthanide products that are needed e.g. for electronics and medical applications.119 Other disadvantages of solvent extraction compared to ion exchange include handling of large solvent volumes as well as producing lots of waste.118, 120 Ion exchange chromatography allows relatively fast separation of the element in question from the target material, which is a critical point for medical radionuclides with short half-lives (e.g. 149Tb with 4 h half-life).

36 Ion exchange chromatography is based on the use of a column packed with ion exchange resin (stationary solid phase). Resins consist of an inert polymer support and functional groups that are covalently bound to the support and contain the ion which can be exchanged. When a solution is passed through the column, ions from the solution can be retained by the resin by the exchange of ions with those of the functional groups. An eluent of choice (mobile phase) can then be passed through the column to allow ions bound to the stationary phase to detach and exit the column. The type of ion exchange resin depends on the mobile phase ions interacting with exchangeable counter ion, associated with the resin functional groups.121

Another chromatographic approach for lanthanide separation is extraction column chromatography. When a solution containing ions is passed through the column packed with extraction resin, ions from the solution can form complexes with functional groups of the resin. These functional groups are often based on existing solvent extraction agents. Similarly to ion exchange chromatography, ions can be detached from the functional groups of the stationary phase by passing a suitable eluent through the column.122

LN resins are extraction resins especially designed for lanthanide separations. Ex- tractants in LN1, LN2, and LN3 resins are bis(2-ethyl-1-hexyl)phosphoric acid (HDEHP), 2-ethyl-1-hexyl(2-ethyl-1-hexyl)phosphonic acid (HEH[EHP]), and bis(2,4,4-trimethyl-1-pentyl)phosphinic acid (H[DTMPP]), respectively123 (Figure 5.1). Behavior of lanthanides on the LN1 and LN2 resins has been studied previously, especially using nitric acid as a mobile phase (Table 5.2).

37 Figure 5.1: Structures of the extractants in LN resins.123

Table 5.2: Lanthanide behavior on LN1 and LN2 extraction resins using nitric acid (or hydrochloric acid) as mobile phase.

Resin Lanthanides Eluent

122 Nd/Pm 0.25 M HNO3 124 Nd/Pm 0.18 M HNO3 122 Sm/Eu/Gd 0.4 M HNO3 125 LN1 Eu/Gd/Tb 0.6 M HNO3 124 Gd/Tb 0.8 M HNO3 124 Dy/Ho 1.4 M HNO3 124 Yb/Lu 3.4 M HNO3 122 Dy/Ho 0.5 M HNO3 122 LN2 Yb/Lu 1.5 M HNO3 126 Yb/Lu 1.5 M HNO3 Yb/Lu126 1.5 M HCl

38 Horwitz et al. reported attempts on several diﬀerent separations of adjacent lanthanides:122 Nd/Pm and Sm/Eu/Gd on LN1 resin and Dy/Ho and Yb/Lu on LN2 resin. In each case, the elution peaks of adjacent lanthanides overlapped so they could not be completely separated under the applied experimental conditions.

In another study126 Horwitz et al. reported Yb/Lu separation on LN2 column, comparing elutions with 1.5 M HNO3 and 1.5 M HCl (Figure 5.2). A slightly better separation was obtained with HNO3 as the eluent but diﬀerences between these experiments were minor.

39 Figure 5.2: Separation of Yb and Lu on LN2 resin: elution with 1.5 M HNO3 (top graph) and 1.5 M HCl (bottom graph).126

40 Monroy-Guzman and Salinas reported separation of the adjacent lanthanides Gd/Tb,

124 Nd/Pm, Dy/Ho, and Yb/Lu on LN1 resin. Diﬀerent eluent (HNO3) concentrations were tested in this study in order to ﬁnd optimal conditions for the separations. Optimal HNO3 concentrations for each separation are presented in Table 5.2.

These separations of adjacent lanthanides suggest that it should be possible to separate the Er from both Ho and Tm using LN resins. As previously mentioned, the objective of the radionuclide development group at Paul Scherrer Institut is to develop production and separation of 165Er via the indirect production route 166Er(p,2n)165Tm → 165Er. Purification of the product would include two steps: firstly, separation of 165Tm from the irradiated target material and secondly, separation after 1 day waiting time109 of the resultant 165Tm to decay to the desired product – 165Er – and final elution thereof. The first separation step is carried out using a cation-exchange resin, as the capacity of the resin to handle ”bulk” quantities is high, as is the separation factor (depending on the eluent). LN extraction resins are not effective for this purpose, but are, however, useful for trace separation and product concentration. Investigating the suitability of LN resins for the second separation step, therefore, is the goal of the present work. Ideally, an LN resin could be used in a generator-like manner, eluting 165Er from the same column several times until all of the 165Tm has decayed. In order to achieve this, different experimental conditions for the separation need to be investigated to find the most favorable conditions and obtain an optimal separation.

Several factors, such as eluent concentration and flow rate, affect the separation efficiency. When eluent concentration is increased, the analyte can be recovered faster from the column, in a smaller eluate volume. Increasing the eluent concentration, however, also makes the separation resolution poorer. The effects of

41 eluent concentration are well demonstrated by Kazakov et al. for the separation of

Tb from Eu and Gd: when HNO3 concentration is increased, peaks in the elution profile become narrower but also move closer together.125 Decreasing eluent flow rate can result in better resolution, since eluent has more time to interact with the resin. Decreasing the eluent flow rate also makes the separation procedure more time consuming so the physical half-life of the nuclides in question need to be considered when choosing suitable flow rate. Column length (amount of resin) also affects the separation. Increasing column length can result in better separation resolution since there is a higher number of functional groups to interact with the analytes. From a shorter column, however, it is possible to obtain the analyte in a smaller eluent volume.

5.4 The concept of distribution coeﬃcients

When assessing the possibilities of separating two or several elements under chosen experimental conditions it is useful to determine distribution coefficients (Kd) of these elements. Distribution coefficient describes the degree to which analyte is retaining on the solid phase (resin of interest) under chosen experimental conditions. High distribution coefficient (>1000) indicates that the element of interest is strongly retained on the resin. When the distribution coefficient is in the

100 < Kd < 1000 interval, the element of interest is moving slowly with the mobile phase. Distribution coeﬃcient value between 0 and 10 represents a situation where an analyte is passing through the resin without any interaction with the solid phase.

Distribution coeﬃcient, Kd is equal to the concentration of an ion of the element in the immobile phase divided by its concentration in the liquid phase according to Equation (5.1):127

42 Ms Vl Kd = · , (5.1) Ml m

where Vl is the volume of the eluent (mL), m is the mass of the resin (g) and

Ms and Ml are the masses of the element (g) in solid and the liquid phases, respectively.

There are two diﬀerent methods to determine distribution coeﬃcient, batch method and column method. In the batch method, resin is added to a solution that contains ions of the element for which Kd is to be determined. The mixture is stirred until equilibrium is reached. Finally resin and solution are separated and concentration (mass) of the ion in both phases is determined. In the column method the resin is packed into a column and solution containing the ions is passed through the column.121

0 Distribution coeﬃcients Kd and Kd of two elements under the same experimental conditions can be used to calculate a separation factor α for these elements. Separation factor is calculated as the ratio between the distribution coeﬃcients according to equation (5.2).

Kd α = 0 , (5.2) Kd

43 The higher the separation factor, the more readily the elements can be separated under the given conditions. Adjacent lanthanides are challenging to separate very eﬀectively.128 However, like distribution coeﬃcients, the values of separation factors depend on the experimental conditions. In previously mentioned study by Horwitz et al.,122 the following separation factors were obtained: 2.0 for Nd/Pm separation on LN1 resin, 1.9 and 1.6 for Sm/Eu and Eu/Gd pairs on LN1 resin, respectively, 2.5 for Dy/Ho separation on LN2 resin and 2.2 for Yb/Lu on LN2 resin.

The possibility of separating elements can also be assessed by determining retention capacities, k, of the elements in question. Retention capacities describe the fraction of the total amount of the element retaining on the resin in given experimental conditions.

6 Aim of Thesis

The aim of this work was to ﬁnd suitable experimental conditions for an appropriate Er/Tm separation on LN resins that would allow Er elution in a small volume of low-concentrated acid in order to facilitate preclinical studies.

Retention capacities for Ho, Er, and Tm on LN1, LN2, and LN3 resins, were determined in order to understand the retention behavior of these lanthanides of interest. The conditions for the determination of the retention capacities (expressed as percentages) were chosen speciﬁcally to simulate a 2-column separation method, developed by the Radionuclide Development Group (Paul Scherrer Insti- tut, Switzerland).

In addition, the column separation experiments were performed on LN2 and LN3 resins to investigate the separation of Er and Tm.

44 The results of this study will be considered while developing 165Er puriﬁcation process which is required after 165Er production via the indirect 166Er(p,2n)165Tm → 165Er nuclear reaction.

Part II: Experimental

7 Materials and methods

7.1 Retention capacities

7.1.1 Sample preparation for determination of retention capacities

In order to understand the retention behavior of Ho, Er and Tm on the LN1 (Triskem International, France), LN2 and LN3 (Eichrom Technologies, USA) extraction resins, retention capacities (expressed as percentage) of the lanthanide of interest were determined. Required masses (200 mg) of each resin were calculated based on the capacities (Table 7.1), stated by the manufacturer, and element concentration to avoid element breakthrough due to the resin saturation.

Table 7.1: Capacities of the LN resins as stated by the manufacturer.

Resin Capacity (mmol/mL) LN1 0.16 LN2 0.16 LN3 0.18

45 Stock solutions (100 ppm) of the lanthanide of interest were prepared by dilution of the 1000 ppm Ho, Er or Tm ICP standards (Sigma-Aldrich, USA) with 0.05 M α-hydroxy-isobutyric acid (α-HIBA, Sigma-Aldrich, USA). The solutions were equilibrated with 200 mg of LN1, LN2 or LN3 resin for 18 hours. Reference samples (without addition of the resin) were prepared under the same experimental conditions. The pH of the stock solutions was adjusted to 4.5 to simulate a previous lanthanide separation method developed by the Radionuclide Development Group (Paul Scherrer Institut, Switzerland). After reaching equilibrium, stock solutions were poured through 25 mL ISOLUTE R Phase separator columns (Bio- tage, Sweden) and collected after passing through the column. These solutions were referred to as the filtrate solutions. Afterwards, hydrochloric acid (Merck, Germany) at different concentrations (0.01 M, 0.05 M, 0.1 M, 0.5 M, 1.0 M and 3.0 M) was passed through the columns and collected (referred to as the eluate solutions). After the experiment, 6.0 M HCl was poured through the column to ensure complete element elution from the resin. Experiments were repeated three times for assessing the reliability of the results. In order to investigate a possible influence of α-HIBA on the retention of the lanthanide of interest on each resin, samples for each resin/lanthanide pair in the absence of α-HIBA were prepared and processed as described before using 3.0 M HCl as eluent for each sample.

7.1.2 Sample preparation for ICP–OES measurements

Concentration of the lanthanide of interest in ﬁltrates and eluates was determined by inductively coupled plasma optical emission spectroscopy (ICP-OES, OPTIMA 3000, Perkin Elmer, USA). The samples for ICP analysis were prepared by dilution of the ﬁltrate and eluate solution aliquots with MilliQ water in order to have 0.3 M HCl concentration in each sample. Calibration standards (1 ppm, 10 ppm, 50

46 ppm, 100 ppm and 150 ppm) were prepared by dilution of the 1000 ppm ICP standards.

7.1.3 Calculations for determination of retention capacities

The measured lanthanide concentrations of the ICP–OES samples are presented in Appendix A. Since the original ﬁltrate and eluent samples were diluted for ICP–OES measurements, the measured concentrations c0 were multiplied by the dilution factor d in order to get the lanthanide concentration c of the original sample (from which aliquots were taken to prepare the ICP–OES samples).

c = d c0 (7.1)

Since the volumes V of the ﬁltrates and eluates were known, masses m of the lanthanides in each sample could be calculated according to equation (7.2):

m = c V (7.2)

The retention capacities could then be calculated according to equation (7.3):

m − m − m k = ref f e · 100 % , (7.3) mref

47 where masses mref , mf , and me are the masses of lanthanide in reference sample, ﬁltrate, and eluate (from the same resin sample as ﬁltrate).

7.2 Column separation experiments

In order to investigate experimental conditions for appropriate Er/Tm separation on LN resins, bench column experiments were performed with the use of radiotrac-

165 171 ers ( Er with t1/2 = 10.36 h and Tm with t1/2 = 1.92 y). In some experiments 166m Ho (t1/2 = 1200 y) was used instead of Er tracer due to the limited availability of the irradiation facilities (IP2 and SINQ at PSI) for the production of Er tracers during the period of the present work.

Column packing solutions of LN3 and LN2 resins were prepared by mixing 1 g

R of the resin in question with 30 mL of 0.01 M HNO3 (Suprapur quality, Merck, Germany) to create a slurry. 1 mL columns (Biotage, Sweden) were packed with the resin slurries and another frit was placed on the upper end of the bed (in addition to the one in the lower end).

The lanthanides stock solution was prepared by mixing 1 mL of 0.05 M α-HIBA with 7.4 µL (74 ng Tm) and 21.4 µL (214 ng Er) of 10 ppm Er (in some experiments – Ho) and Tm ICP standards, respectively. Required masses of Ho/Er and Tm were calculated to imitate the presence of 5 GBq of 165Er and 5 GBq of 165Tm in the system. Aliquots of 166mHo/165Er and 171Tm tracers were added for mon- itoring Ho, Er and Tm behavior during the separation process. Activities of the radiotracers were measured using high-purity germanium (HPGe) detector (Can- berra, France) in combination with the InterWinner software package (version 7.1, Itech Instruments, France).

At the beginning of each experiment, resin was cleaned by rinsing with 5 mL of

48 2.0 M HNO3 and 15 mL of MilliQ water. After loading the lanthanides stock solution, the column was washed with 20 mL of MilliQ water in order to remove possible trace amounts of α-HIBA. Elution was carried out using HCl of diﬀerent concentrations (0.01 M – 2.0 M). Flow rate of 0.3 mL/min or 0.6 mL/min was applied for the separation process and 5 mL fractions were collected. To determine the presence of the tracers and their activities, collected fractions were measured by gamma spectrometry. Based on these data, elution proﬁles were obtained. The activity of 166mHo was determined based on its several γ-transitions (81 keV, 184 keV, 280 keV, 411 keV, 530 keV, 571 keV, 671 keV, 712 keV, 752 keV, 810 keV, 831 keV),68 the activity of 165Er was determined based on X-ray emissions at the energies 46.7 keV (21.6 %) and 47.5 keV (38.4 %)68 and the activity of 171Tm was determined based on X-ray emissions at 51.4 keV (0.3 %) and 52.4 keV (0.5 %).68 Since the intensity of 165Er X-rays (21.6 % and 38.4 %) is very large compared to that of 171Tm (0.3 % and 0.5 %), it was challenging to detect Tm in the fractions containing also Er. In order to detect Tm accurately, some fractions were re-measured after 165Er had decayed.

49 8 Results and discussion

8.1 Retention capacities

Retention capacities (expressed as percentage) were calculated from the masses of lanthanides in diﬀerent samples obtained from the ICP–OES measurements using equation (7.3). Calculated values for each experiment are presented in Tables 8.1–8.4. For easier comparison of percentage values determined for diﬀerent lanthanides, average values (Table 8.5) were calculated based on the three separate experiments. Also separation factors on each LN resin were calculated for Ho/Er and Er/Tm lanthanide pairs using equation (5.2). Average values of separation factors are presented in Table 8.6.

Table 8.1: Retention capacities (in %) of the lanthanides of interest in HCl media using LN1 resin.

Element Exp. 3 M 1 M 0.5 M 0.1 M 0.05 M 0.01 M 1 62 79 92 100 100 100 Ho 2 70 80 93 100 100 100 3 51 80 92 100 100 100 1 69 84 96 100 100 100 Er 2 62 82 95 100 100 100 3 70 85 96 100 100 100 1 59 83 98 100 100 100 Tm 2 65 85 97 100 100 100 3 80 92 98 100 100 100

50 Table 8.2: Retention capacities (in %) of the lanthanides of interest in HCl media using LN2 resin.

Element Exp. 3 M 1 M 0.5 M 0.1 M 0.05 M 0.01 M 1 9 11 15 95 100 100 Ho 2 4 7 5 95 99 100 3 13 12 17 95 99 100 1 9 12 36 99 100 100 Er 2 8 8 14 97 100 100 3 4 5 15 98 100 100 1 13 9 38 99 100 100 Tm 2 2 3 29 99 100 100 3 8 14 44 99 100 100

Table 8.3: Retention capacities (in %) of the lanthanides of interest in HCl media using LN3 resin.

Element Exp. 3 M 1 M 0.5 M 0.1 M 0.05 M 0.01 M 1 4 43 41 59 72 98 Ho 2 5 7 7 5 7 97 3 13 17 16 16 22 98 1 3 37 43 59 69 99 Er 2 1 6 6 4 17 98 3 3 4 3 5 18 99 1 5 6 8 8 37 100 Tm 2 8 9 8 7 41 100 3 6 10 9 12 39 99

51 Table 8.4: Retention capacities (in %) of the lanthanides of interest in 3 M HCl on LN1, LN2, and LN3 resins, determined for samples in the absence of α-HIBA.

Element LN1 LN2 LN3 Ho 8 8 29 Er 14 7 17 Tm 18 7 12

Table 8.5: Average retention capacities (in %) calculated from Tables 8.1–8.3.

Resin Element 3 M 1 M 0.5 M 0.1 M 0.05 M 0.01 M Ho 61 ± 8 79 ± 0 92 ± 0 100 ± 0 100 ± 0 100 ± 0 LN1 Er 67 ± 4 84 ± 1 96 ± 0 100 ± 0 100 ± 0 100 ± 0 Tm 68 ± 9 86 ± 4 98 ± 1 100 ± 0 100 ± 0 100 ± 0 Ho 9 ± 3 10 ± 2 12 ± 5 95 ± 0 99 ± 0 100 ± 0 LN2 Er 7 ± 2 8 ± 3 22 ± 10 98 ± 1 100 ± 0 100 ± 0 Tm 8 ± 4 9 ± 4 37 ± 6 99 ± 0 100 ± 0 100 ± 0 Ho 7 ± 4 22 ± 15 22 ± 14 26 ± 23 34 ± 28 98 ± 1 LN3 Er 2 ± 1 15 ± 15 17 ± 18 23 ± 26 35 ± 24 99 ± 0 Tm 6 ± 1 8 ± 2 9 ± 1 9 ± 2 39 ± 2 100 ± 0

52 Table 8.6: Average separation factors of Ho/Er and Er/Tm lanthanide pairs on LN resins.

Resin Ln pair 3 M 1 M 0.5 M 0.1 M 0.05 M 0.01 M LN1 Ho/Er 1.2 ± 0.2 1.1 ± 0.0 1.0 ± 0.0 1.0 ± 0.0 1.0 ± 0.0 1.0 ± 0.0 Er/Tm 1.0 ± 0.1 1.0 ± 0.0 1.0 ± 0.0 1.0 ± 0.0 1.0 ± 0.0 1.0 ± 0.0 LN2 Ho/Er 1.1 ± 0.8 0.9 ± 0.3 2.0 ± 0.8 1.0 ± 0.0 1.0 ± 0.0 1.0 ± 0.0 Er/Tm 1.2 ± 0.7 1.3 ± 1.1 2.0 ± 0.8 1.0 ± 0.0 1.0 ± 0.0 1.0 ± 0.0 LN3 Ho/Er 0.4 ± 0.3 0.7 ± 0.3 0.7 ± 0.4 0.7 ± 0.3 1.4 ± 0.7 1.0 ± 0.0 Er/Tm 3.8 ± 3.0 1.4 ± 1.0 1.5 ± 1.2 1.4 ± 1.0 1.7 ± 0.8 1.0 ± 0.0

Under the same experimental conditions retention percentages for the same resin/ lanthanide pair varied considerably for three independent repetitions (Tables 8.1– 8.3). Also, when comparing behavior of Ho, Er and Tm on the same resin, the expected decrease in the retention from Ho to Er and to Tm122–124 was not always observed. These differences could be due to the following reasons. The ICP– OES device could provide slightly variable element concentrations on the different measurement times. The variation in measured values could be detected when comparing background signals obtained in different experiments. The background was in the range -0.4 ppm – 0.8 ppm depending on the measurement date. Even though the measurement system was always rinsed in between sample measurements, in some experiments, additional blank samples (containing media without analyte) were measured in order to rinse the system more after measuring samples of high lanthanide concentration. The blank sample concentration, in these cases, was repeatedly ca. 0.3 ppm higher than the background (8 ppm higher in one example). Also, two tests were performed in which the same 50 ppm sample (prepared from 1000 ppm ICP standard) was measured two consecutive times and in both tests the difference between the measured values was 6 ppm. Taking into

53 account that all measured sample concentrations were relatively low (<50 ppm, in most cases <10 ppm), even small errors in the obtained values could have a rather significant effect on the final retention capacities, calculated based on the ICP– OES measurement results. Human error (e.g. pipetting) should also be considered. Pipetting errors could have occured when pipetting ICP standard solutions to the weighed resin samples or when taking aliquots of filtrates and eluates to prepare the samples for ICP–OES measurements.

Another possible reason for the variation in measured element concentrations can originate from the lanthanide elution step. Before the resin/lanthanide samples were poured into the columns, samples were equilibrated for 18 hours. With the elution step, eluent was passed through the column without resident time on the resin. There was a significant variation in the eluent flow rates, depending on the column (some columns allowed the eluent pass through very fast, whereas others seemed to be somewhat blocked), meaning that it was not possible to control the interaction time between eluent and resin. Differences in interaction times between different samples were the order of minutes. Since it takes several hours to reach equilibrium, differences of this magnitude should not cause any considerable error to the results. Especially, for eluent concentrations on which the retention of the lanthanide on the resin is very poor or very efficient, the interaction time should not affect the retention behavior. The possibility of errors of this nature, however, should not be completely disregarded for samples that were eluted with HCl concentration in between the two extremities.

Despite these uncertainties, for the same lanthanide on LN1, LN2 and LN3 resins a similar trend was observed: when eluent concentration was decreased, the lanthanide of interest was more strongly held by the resin. For example, at 1.0 M or 3.0 M HCl using LN1 resin, Ho moves with the mobile phase (based on the

54 determined retention capacities), but is strongly retained when using eluent concentration <0.5 M.

Behavior of both Er and Tm on LN1 is quite similar to that of Ho, although retention percentages obtained for Er, as well as Tm, are on average higher than those for Ho (Table 8.5), when the lanthanides are not completely retained. When comparing the values of retention percentages of single experiments (Table 8.1), it can be seen that the results for samples eluted with 3.0 M HCl have the greatest variation (e.g. values for Ho, highlighted in pink). Also for Tm, retention percentages are significantly higher in one experiment compared to the other two (values highlighted in yellow in Table 8.1). Differences between average values for different lanthanides (Table 8.5), for any given HCl concentration are, however, very small, indicating similar behavior of Ho, Er and Tm. Hence, according to the results obtained in these experiments one could conclude that it may be challenging to perform separation of the lanthanides of interest on LN1 resin in HCl medium. However, by varying column length and flow rate it should be possible to find experimental conditions that allow sufficient separation.

Results on LN2 resin (Table 8.2) for all studied lanthanides clearly suggest that the analyte moves through the column with eluent concentrations ≥0.5 M, whereas for concentrations ≤0.1 M the analyte is strongly retained by the resin. There are no consistent differences between retention percentages for Ho, Er, and Tm on LN2 resin. In the averaged values (Table 8.5), however, there is a difference in the retention percentages for samples eluted with 0.5 M HCl. The difference for this eluent concentration is also seen in separation factors calculated for Ho/Er and Er/Tm pairs on LN2 resin (Table 8.6). For both lanthanide pairs separation factors are close to 1, except for HCl concentration 0.5 M. With this concentration, separation factor is 2.0 ± 0.8 for both Ho/Er and Er/Tm pair. Separation factors

55 with each eluent concentration are, however, close to 1. This suggests that the lanthanides of interest all behave similarly on LN2 resin and hence, it is very unlikely that Ho/Er and Er/Tm separations could be performed on LN2 resin in HCl medium.

There are some discrepancies in the results of individual experiments. For both Ho and Tm, retention percentages obtained in the second experiment are considerably lower (highlighted in Table 8.2 in green and orange for Ho and Tm, respectively) compared to the other two experiments. For Er, only one single value (highlighted in blue in Table 8.2) is more than two fold compared to the other two values determined in the repetition experiments. However, averaged values (Table 8.5) for the experiments on LN2 resin followed the observed trend: values are increasing when eluent concentration decreases and also increasing according to the atomic number of lanthanide when compared between elements.

On LN3 resin Ho, Er, and Tm move with the liquid phase with eluent concentration 0.05 M and higher, retaining strongly on the resin only when applying the lowest of the studied eluent concentrations (0.01 M). Differences in four individual experiments were observed for LN3 resin (Table 8.3). For Ho, there is considerable variation in the values of retention percentages obtained in all three repetitions (values highlighted in purple in Table 8.3). The largest differences for corresponding values are more than tenfold. For both Ho and Er, retention percentages determined in the first experiments (values for Er highlighted in cyan in Table 8.3) are significantly higher compared to values determined in the other experiments. Unlike for LN1 and LN2, the average retention percentages for LN3 resin at high HCl concentrations (values highlighted in blue in Table 8.5) do not follow the expected trend, when compared between the three lanthanides. On the contrary, values seem to diminish in the order Ho → Er → Tm. Partly this is due to the

56 exceptionally high values obtained for Ho and Er in the ﬁrst experiments.

In general, based on these experiments, it can be stated that retention percentages of the lanthanide of interest on LN resins under the same experimental conditions decrease in order LN1>LN2>LN3 which is in agreement with literature. This is due to the decrease in the acidity of the extractant in the resin, being the most acidic in LN1 resin (LN1>LN2>LN3).123

8.2 Column separation experiments

Based on the determined retention percentages, LN3 resin was chosen for further experiments as it allows elution of the lanthanide of interest with lower eluent concentration as compared to LN1 and LN2 resins. Since for radiopharmaceutical applications the radionuclide precursor is preferred in small volume of diluted acid, one can conclude that it is most logical to test separations of the investigated lanthanides primarily on LN3 resin.

8.2.1 Column separation of Ho and Tm

Based on the retention percentages determined for Ho and Tm on LN3, it can be estimated that both of these elements start moving with the mobile phase on the resin between eluent concentrations 0.01 M and 0.05 M. With concentration of 0.01 M Ho and Tm are strongly retained on the resin (retention percentages ≥ 98 %), whereas with eluent concentration of 0.05 M retention percentages are considerably lower on average, (34 ± 28) % for Ho and (39 ± 2) % for Tm.

Three Ho/Tm separations were performed on LN3 resin column. The results are shown in Figures 8.1 – 8.3. From 46 mm x 6 mm column, Ho was eluted with 0.04 M HCl, whereas Tm began to come out from the column when 0.05 M eluent

57 concentration was applied (Figure 8.1). A sufficient separation was achieved: 99 % of Ho was eluted without Tm present in the fractions. Ho was eluted in total volume of ca. 50 mL of eluent, which is a significantly larger volume than would be desired for the application in nuclear medicine. The separation process took 3 hours with the use of 0.3 mL/min flow rate. Separation should be optimized in a way that would allow it to be carried out faster. However, in this case, it is not crucial since the separation method would be applied in a situation where there is ingrowth of 165Er from 165Tm.

Figure 8.1: Elution curve of Ho/Tm separation on LN3 resin (46 mm x 6 mm) as a function of eluent volume and HCl concentration. Eluent concentrations (top axis): 0.01 M – 2 M HCl; ﬂow rate: 0.3 mL/min.

In order to decrease separation time and eluent volume in which Ho is obtained, the ﬂow rate was increased to 0.6 mL/min while leaving the same column length (46 mm x 6 mm). The starting HCl concentration was chosen to be 0.04 M based on the results from the previous experiment. As can be seen from Figure 8.2, Ho

58 began to elute after 10 mL HCl volume and 68 % of Ho was eluted in the total volume of 124 mL HCl (90 mL of 0.04 M HCl and 34 mL of 0.05 M HCl) before Tm was detected in the first Ho fraction. The amount of Ho was smaller compared to the previous experiment as a result of increased tailing due to the change in flow rate. The tailing of the Ho peak also resulted in a shorter gap between Ho and Tm peaks. These differences can be further explained by the difference in the applied eluent concentrations: in the previous experiment Ho could have already moved in the column when HCl concentrations of 0.1 M – 0.3 M were applied. Compared to the previous experiment, the separation time was shortened to ca. half (1.5 h) due to the change in flow rate and eluent concentration.

Figure 8.2: Elution curve of Ho/Tm separation on LN3 resin (46 mm x 6 mm) as a function of eluent volume and concentration. Eluent concentrations (top axis):

0.04 M HCl, 0.05 M HCl, and 2.0 M HNO3; ﬂow rate: 0.6 mL/min.

In order to test the eﬀect of the column length on the required eluent volume, a 15 mm LN3 resin was tested (Figure 8.3). Despite the increase in tailing of the Ho

59 peak, it was concluded that 0.6 mL/min flow rate should be used instead of 0.3 mL/min as it allowed obtaining efficient Ho/Tm separation within 1.5 hours (3 hours at 0.3 mL/min). Under these experimental conditions, 87 % of Ho was eluted in 54 mL of 0.04 M HCl before Tm was first detected. The gap between Ho and Tm becomes considerably closer. It was concluded that, while the shorter column length resulted in a smaller eluent volume required, the gap obtained between Ho and Tm elution was too close, since the optimized separation conditions should be applied to Er/Tm separation and the gap between Er and Tm peaks was expected to be even closer than the gap between Ho and Tm.

Figure 8.3: Elution curve of Ho/Tm separation on LN3 resin (15 mm x 6 mm) as a function of eluent volume and concentration. Eluent concentrations (top axis):

0.04 M HCl, 0.05 M HCl, and 2.0 M HNO3; ﬂow rate: 0.6 mL/min.

60 8.2.2 Column separation of Er and Tm

In order to investigate experimental conditions for Er/Tm separation on LN3 resin, four column experiments were performed with Er and Tm tracers. The results are shown in Figures 8.4 – 8.7. 165Er tracers were produced with Injector 2 cyclotron at Paul Scherrer Institute via 165Ho(p,n)165Er reaction. After production, 165Er was separated from the irradiated Ho target by Radionuclide Development Group at PSI. Due to the relatively short half-life of 165Er, Er/Tm column separation experiments were performed directly after 165Er puriﬁcation.

According to the previous results, Ho could be eﬃciently separated from Tm on LN3 resin column with the use of 0.04 M HCl at a ﬂow rate of 0.6 mL/min. These experimental conditions were afterwards applied for Er/Tm separation. In the

166 68 first experiment, Ho (t1/2 = 26.8 h) was not added intentionally (Figure 8.4), however, the presence of 166Ho as a tracer impurity allowed more information to be obtained about its behavior on LN3 resin column. Ho started to be eluted after ca. 10 mL volume of 0.04 M HCl and Er after 30 mL eluent volume. Significant over- lapping of Ho and Er peaks was observed, indicating that an effective separation of these elements was not possible under the chosen experimental conditions. 165Er fractions obtained from the performed Er/Tm separation may also contain 165Ho, since it is the daughter nuclide of 165Er. Ho being present as an impurity can lower Er labeling yield. The elution profile of Ho is in good agreement with the elution profile obtained before (Figure 8.2). Elution was stopped before eluting Tm with

2.0 M HNO3 in order to immediately load collected Er fraction on LN2 resin (3 mm x 6 mm) and investigate whether LN2 resin column could be applied for the product concentration step. The results of the concentration experiment are presented in section 8.2.3. Tm that was left in the LN3 column was only eluted from the column after several days. This explains why there is no active Ho observed

61 along with Tm, when eluting with 2.0 M HNO3, like in the previous experiments (Figures 8.2 and 8.3) . For the same reason, Er tailing that could be expected due the similar behavior of Ho on LN3 resin, seems to stop abruptly.

Figure 8.4: Elution curve of Er and Ho on LN3 resin (46 mm x 6 mm) as a function of eluent volume and concentration. Eluent concentration (top axis): 0.04 M HCl; ﬂow rate: 0.6 mL/min.

In order to see the complete elution behavior of Er and possible overlap between Er and Tm, the separation experiment was repeated using the same starting concentration of HCl, as well as the same column and eluent ﬂow rate (Figure 8.5). In total, 93 % of all Er could be eluted before any Tm was detected, also being eluted with 0.04 M HCl, as previously observed when using the shorter LN3 column (Figure 8.3). Hence, these two elements cannot be completely separated under the chosen conditions but suﬃcient amount of Er can still be obtained well before Tm starts to be eluted. 90 % of Er was obtained in a volume of 125 mL of 0.04 M HCl.

62 Figure 8.5: Elution curve of Er/Tm separation on LN3 resin (46 mm x 6 mm) as a function of eluent volume and concentration. Eluent concentrations (top axis):

0.04 M, 0.05 M, and 0.1 M HCl, and 2.0 M HNO3; ﬂow rate: 0.6 mL/min.

In order to assess the reliability of the results, the separation was repeated under the same experimental conditions (Figure 8.6). When comparing the elution curves of both experiments (Figures 8.5 and 8.6) it can be concluded that the results are in good agreement. In the repetition experiment, even 98 % of Er was eluted before Tm began to be eluted and 90 % of Er (without any Tm) could be obtained in 100 mL of 0.04 M HCl compared to the 125 mL in the previous experiment. However, based on these results, it can be stated that Er cannot be eluted in a small volume of HCl in the experimental conditions in question.

63 Figure 8.6: Elution curve of Er/Tm separation on LN3 resin (46 mm x 6 mm) as a function of eluent volume and concentration. Eluent concentrations (top axis):

0.04 M, 0.05 M, and 0.1 M HCl, and 2.0 M HNO3; ﬂow rate: 0.6 mL/min.

Er and Tm can be separated on 46 mm long LN3 resin, however, there is no gap between the elution peaks. In order to be able to separate Er and Tm more efficiently, the eluent concentration was decreased to 0.03 M while leaving flow rate the same (0.6 mL/min) (Figure 8.7). According to the elution profile, the resolution was improved: the gap between Er main fraction and first Tm fraction is ca. 50 mL. However, the required eluent volume becomes significantly larger. 98 % of Er could be obtained without any Tm present but the total eluent volume required was 315 mL. 90 % of Er could be obtained in ca. 215 mL of 0.03 HCl but this volume is too large, even much larger than the corresponding volumes in the previous experiments (125 mL and 100 mL). Also along with the larger eluate volume, elution time becomes longer: 215 mL with 0.6 mL/min flow rate corresponds to elution time of 6 hours. Hence, it can be concluded that Er/Tm

64 separation using 0.04 M HCl as eluent is a better option than elution with 0.03 M HCl.

Figure 8.7: Elution curve of Er/Tm separation on LN3 resin (46 mm x 6 mm) as a function of eluent volume and concentration. Eluent concentrations (top axis):

0.03 M HCl and 2.0 M HNO3; ﬂow rate: 0.6 mL/min.

In general, in each experiment Tm started to be eluted with the same eluent concentration as Er. While Er and Tm can be separated using LN3 resin, the eluent volume required to elute a suﬃcient fraction of the total Er activity is too large.

65 8.2.3 Er concentration experiment

As 165Er is eluted from the 46 mm and 15 mm long LN3 resin columns in a large eluent volume, it should be further concentrated to facilitate preclinical studies. Suitability of small LN2 column was investigated for this purpose based on the retention capacities (Tables 3 and 6). As from the LN3 resin, 165Er is eluted in 0.04 M HCl, it should be possible to directly load it to the short LN2 resin column with 100 % retention (Table 3) and then elute it in a smaller volume of the eluent. According to the retention capacities determined for Er in LN2 resin, Er should begin to move in the column with eluent concentration between 0.1 M and 0.5 M. This was tested with a 3 mm long LN2 column and the elution profile of the experiment is presented in Figure 8.8. When loading 0.04 M HCl solution containing the Er into the LN2 column breakthrough of 6 % was detected. In order to obtain Er in as diluted HCl as possible, HCl concentration of 0.1 M was applied first. With 0.1 M eluent concentration, Er retained on the resin as expected. When applying eluent concentration 0.2 M, Er began to elute slowly. After applying concentrations 0.4 M and 0.5 M, Er eluted efficiently from the column. Most of the Er (94 %) was eluted in ca. 5 mL of HCl, however, this volume is too high. The required eluent volume could possibly be slightly decreased by eluting directly with 0.5 M HCl. It appears that LN2 resin cannot be utilized to concentrate Er in a volume ≤1 mL using 0.1–0.5 M HCl to facilitate preclinical studies.

66 Figure 8.8: Elution curve of Er on LN2 resin (3 mm x 6 mm) as a function of eluent volume and concentration. Eluent concentrations (top axis): 0.1 – 2.0 M HCl; ﬂow rate: 0.6 mL/min.

Part III: Conclusions and outlook

Retention percentages obtained for the lanthanides of interest on diﬀerent LN Resins follow the expected pattern: under the same experimental conditions retention percentages decrease in order LN1>LN2>LN3. The results are in agreement with literature and data provided by the manufacturers. The retention percentages obtained, could be used for choosing suitable conditions for column experiments.

From the column experiments, it can be concluded that it is possible to separate Er

67 and Tm using LN3 resin with 0.04 M HCl as eluent. Decreasing the concentration of the eluent (0.03 M HCl) improves the separation but increases the eluent volume needed to elute the main Er fraction. Based on Ho/Tm separation experiments, decreasing the eluent ﬂow rate also improves the separation. However, decreasing the ﬂow rate makes the separation process more time consuming. Decreasing the length of the LN3 column decreases the volume in which Er is eluted but shortens the gap between Er and Tm. Based on the results of the column experiments it can be concluded that LN3 resin is not suitable for the intended purpose, since eluate volumes are too high and it cannot be applied in generator-like way due to Tm also moving on the resin with similar eluent concentrations as Er.

LN3 resin could be applied for Er/Tm separation in cases where higher required eluent volume is acceptable. In order to use Er obtained in this way for radiopharmaceutical purposes, an additional concentration step would be necessary.

68 References

[1] Cancer Research UK. What is cancer? https://www.cancerresearchuk.org/about-cancer/what-is-cancer.

[2] L. A. Torre, F. Bray, R. L. Siegel, J. Ferlay, J. Lortet-Tieulent, and A. Jemal. Global cancer statistics. CA-Cancer J. Clin., 15(65):87 – 108, 2015.

[3] National Cancer Institute. Types of cancer treatment, https://www.cancer.gov/about-cancer/treatment/types.

[4] R. Siegel. Cancer treatment and survivorship statistics, 2012. CA-Cancer J. Clin., 62(4):220, 241 2012.

[5] B. W. Stewart and C. P. Wild, editors. World Cancer Report 2014. Inter- national Agency for Research on Cancer, 2014.

[6] Cancer Research UK. How chemotherapy works, https://www.cancerresearchuk.org/about-cancer/cancer-in- general/treatment/chemotherapy/how-chemotherapy-works.

[7] E. Van Limbergen, M. Joiner, A. Van der Kogel, and W. Dörr. Radiobiol- ogy of LDR, HDR, PDR and VLDR brachytherapy. In E. Van Limbergen, R. Pötter, P. Hoskin, and D. Baltas, editors, The GEC ESTRO Handbook of Brachytherapy, chapter 5. ESTRO Radiotherapy & Oncology, 2015.

[8] S. V. Gudkov, N. Y. Shilyagina, V. A. Vodeneev, and A. V. Zvyagin. Tar- geted radionuclide therapy of human tumors. Int. J. Mol. Sci., 17(33):1 – 19, 2016.

[9] G. B. Saha. Fundamentals of nuclear pharmacy. Springer, 7th edition, 2018.

[10] A. Dash, F. F. Knapp Jr., and M. R. A. Pillai. Targeted radionuclide therapy - an overview. Curr. Radiopharm., 6(3):1 – 29, 2013. 69 [11] S. Mathur and J. Sutton. Personalized medicine could transform healthcare. Biomed. Rep., 7:3 – 5, 2017.

[12] M. A. Hamburg and F. S. Collins. The path to personalized medicine. N. Engl. J. Med., 363(4):301 – 304, 2010.

[13] R. M. Subramaniam. Nuclear medicine and molecular imaging - an impactful decade of contributions to patient care and driving precision medicine. Am. J. Roentgenol., 209:241 – 242, 2017.

[14] K.-H. Jung and K.-H. Lee. Molecular imaging in the era of personalized medicine. J. Pathol. Transl. Med., 49:5 – 12, 2015.

[15] S. Carlsson. A glance at the history of nuclear medicine. Acta Oncol., 34(8):1095 – 1102, 1995.

[16] G. Hevesy. The absorption and translocation of lead by plants. Biochem. J., 17(4 – 5):439 – 445, 1923.

[17] F. F. Knapp and A. Dash. Radiopharmaceuticals for therapy. Springer, 2016.

[18] P. J. Gilmer. In M.-H. Chiu, P. J. Gilmer, and D. F. Treagust, editors, Cel- ebrating the 100th Anniversary of Madame Marie Sklodowska Curie’s Nobel Prize in Chemistry, chapter 2. Sense Publishers, 2011.

[19] C. D. Anderson. The apparent existence of easily deﬂectable positives. Sci- ence, 76(1967):238 – 239, 1932.

[20] O. Chiewitz and G. Hevesy. Radioactive indicators in the study of phospho- rus metabolism in rats. Nature, 136(3445):754 – 755, 1935.

[21] N. Kumar, R. K. R. Singh, D. Dutta, and S. C. Kheruka. Lu-177 – a noble tracer: Future of personalized radionuclide therapy. Clinics in Oncology, 2:1 – 3, 2017.

70 [22] S. Bhattacharyya and M. Dixit. Metallic radionuclides in the development of diagnostic and therapeutic radiopharmaceuticals. Dalton Trans., 40(23):6112 – 6128, 2011.

[23] E. B. Silberstein. The snm practice guideline for therapy of thyroid disease with 131I. J. Nucl. Med., 53(10):1 – 19, 2012.

[24] M. R. Harrison, T. Z. Wong, A. J. Armstrong, and D. J. George. Radium-223 chloride: a potential new treatment for castration-resistant prostate cancer patients with metastatic bone disease. Cancer. Manag. Res., 5:1 – 14, 2013.

[25] J. W. Fletcher. Recommendations on the use of 18F-FDG pet in oncology. J. Nucl. Med., 49(3):480 – 508, 2008.

[26] A. Mojtahedi, S. Thamake, I. Tworowska, D. Ranganathan, and E. S. Del- passand. The value of 68Ga-DOTATATE PET/CT in diagnosis and manage- ment of neuroendocrine tumors compared to current FDA approved imaging modalities: a review of literature. Am. J. Nucl. Med. Mol. Imaging, 4(5):426 – 434, 2014.

[27] E. Bombardieri, V. Ambrosini, C. Aktolun, R. P. Baum, A. Bishof-Delaloye, S. Del Vecchio, L. Maﬃoli, L. Mortelmans, W. Oyen, G. Pepe, and A. Chiti. 111In-pentetreotide scintigraphy: procedure guidelines for tumour imaging. Eur. J. Nucl. Med. Mol. Imaging, 37:1441 – 1448, 2010.

[28] S. Banerjee, M. R. A. Pillai, and F. F. Knapp. Lutetium-177 therapeutic radiopharmaceuticals: Linking chemistry, radiochemistry, and practical applications. Chem. Rev., 115:2934 – 2974, 2015.

[29] C. L. Maini, S. Bergomi, L. Romano, and R. Sciuto. 153Sm-EDTMP for bone pain palliation in skeletal metastases. Eur. J. Nucl. Med. Mol. Imaging, 31(Suppl. 1):S171 – S178, 2004.

71 [30] S. P. Nielsen, M. L. Møller, and J. Trap-Jensen. 99mTc-DTPA scintillation- camera renography: A new method for estimation of single-kidney function. J. Nucl. Med., 18(2):112 – 117, 1977.

[31] A. J. Davies. Radioimmunotherapy for B-cell lymphoma: Y90 ibritumomab tiuxetan and I131 tositumomab. Oncogene, 26:3614 – 3628, 2007.

[32] S. R. Cherry, J. A. Sorenson, and M. E. Phelps. Physics in Nuclear Medicine. Elsevier, 2012.

[33] Lumen Physics. Medical imaging and diagnostics, https://courses.lumenlearning.com/physics/chapter/32-1-medical-imaging- and-diagnostics/.

[34] Y. D’Asseler, S. Vandenberghe, F. De. Winter, R. Van de Walle, M. Koole, L. Bouwens, I. Lemahieu, and R.A. Dierckx. PET imaging using gamma cameras. Comput. Med. Imaging Graph., 25:87 – 96, 2001.

[35] J. Langner. Development of a parallel computing optimized head move- ment correction method in positron-emission-tomography. Master’s thesis, University of Applied Sciences, Dresden, 2003.

[36] D. L. Bailey and K. P. Willowson. Quantitative SPECT/CT: SPECT joins PET as a quantitative imaging modality. Eur. J. Nucl. Med. Mol. Imaging, 41(Suppl 1):S17 – S25, 2014.

[37] R. J. Hicks and M. S. Hofman. Is there still a role for SPECT–CT in oncology in the PET–CT era? Nat. Rev. Clin. Oncol., 9:712 – 720, 2012.

[38] C. F. Ramogida and C. Orvig. Tumour targeting with radiometals for diagnosis and therapy. Chem. Commun., 49:4720 – 4739, 2013.

72 [39] J. D. Kelly, A. M. Forster, B. Higley, C. M. Archer, F. S. Booker, L. R. Can- ning, K. W. Chiu, B. Edwards, H. K. Gill, M. McPartlin, K. R. Nagle, I. A. Latham, R. D. Pickett, A. E. Storey, and P. M. Webbon. Technetium-99m- tetrofosmin as a new radiopharmaceutical for myocardial perfusion imaging. J. Nucl. Med., 34(2):362 – 373, 1993.

[40] I. C. Dormehl, D. W. Oliver, K.-J. Langen, N. Hugo, and S. A. Croft. Technetium-99m-HMPAO, technetium-99m-ECD and iodine-123-IMP cere- bral blood ﬂow measurements with pharmacological interventions in pri- mates. J. Nucl. Med., 38(12):1897 – 1901, 1997.

[41] S. Lahiri, M. Maiti, and K. Ghosh. Production and separation of 111In: an important radionuclide in life sciences: a mini review. J. Radioanal. Nucl. Chem., 297:309 – 318, 2013.

[42] J. E. Maxwell and J. R. Howe. Imaging in neuroendocrine tumors: an update for the clinician. Int. J. Endocr. Oncol., 2(2):159 – 168, 2015.

[43] J. B. Bomanji, W. Wong, M. N. Gaze, A. Cassoni, W. Waddington, J. Solano, and P. J. Ell. Treatment of neuroendocrine tumours in adults with 131I-MIBG therapy. Clin. Oncol., 15(4):193 – 198, 2003.

[44] C. Müller, E. Fischer, M. Behe, U. Köster, H. Dorrer, J. Reber, S. Haller, S. Cohrs, A. Blanc, J. Grünberg, M. Bunka, K. Zhernosekov, N. van der Meulen, K. Johnston, A. Türler, and R. Schibli. Future prospects for SPECT imaging using the radiolanthanide terbium-155 – production and preclinical evaluation in tumor-bearing mice. Nucl. Med. Biol., 41:e58 – e65, 2014.

[45] S. L. Pimlott and A. Sutherland. Molecular tracers for the PET and SPECT imaging of disease. Chem. Soc. Rev., 40(1):149 – 162, 2011.

[46] H. Jadvar and J. A. Parker. Clinical PET and PET/CT. Springer, 2005.

73 [47] H. R. Schelbert, M. E. Phelps, S.-C. Huang, N. S. MacDonald, H. Hansen, C. Selin, and D. E. Kuhl. N-13 ammonia as an indicator of myocardial blood ﬂow. Circulation, 63(6):1259 – 1272, 1981.

[48] J. Yang, Y. Kan, B. H. Ge L. Yuan, C. Li, and W. Zhao. Diagnostic role of Gallium-68 DOTATOC and Gallium-68 DOTATATE PET in patients with neuroendocrine tumors: a meta-analysis. Acta Radiol., 55(4):389 – 398, 2014.

[49] A. Pfeifer. 64Cu-DOTATATE PET for neuroendocrine tumors: A prospective head-to-head comparison with 111In-DTPA-Octreotide in 112 patients. J. Nucl. Med., 56(6):847 – 854, 2015.

[50] C. B. Johnbeck. Head-to-head comparison of 64Cu-DOTATATE and 68Ga- DOTATOC PET/CT: A prospective study of 59 patients with neuroendocrine tumors. J. Nucl. Med., 58(3):451 – 457, 2017.

[51] A. Boschi, P. Martini, E. Janevik-Ivanovska, and A. Duatti. The emerging role of copper-64 radiopharmaceuticals as cancer theranostics. Drug Discov. Today, 23(8):1489 – 1501, 2018.

[52] C. Müller, C. Vermeulen, K. Johnston, U. Köster, R. Schmid, A. Türler, and N. P. van der Meulen. Preclinical in vivo application of 152Tb-DOTANOC: a radiolanthanide for PET imaging. EJNMMI Res., 6:35 – 44, 2016.

[53] R. B. Baum, A. Singh, M. Benešová, C. Vermeulen, S. Gnesin, U. Köster, K. Johnston, D. Müller, S. Senftleben, H. R. Kulkarni, A. Türler, R. Schi- bli, J. O. Prior, N. P. van der Meulen, and C. Müller. Clinical evaluation of the radiolanthanide terbium-152: ﬁrst-in-human PET/CT with 152Tb- DOTATOC. Dalton Trans., 46(42):14638 – 14646, 2017.

[54] H. Nagasawa and J. B. Little. Induction of sister chromatid exchanges by extremely low doses of α-particles. Cancer Res., 52:6394 – 6396, 1992.

74 [55] C. Mothersill and C. Seymour. Radiation-induced bystander eﬀects: Past history and future directions. Radiat. Res., 155(6):759 – 767, 2001.

[56] W. F. Morgan. Non-targeted and delayed eﬀects of exposure to ionizing radiation: I. radiation-induced genomic instability and bystander eﬀects In Vitro. Radiat. Res., 159(5):567 – 580, 2003.

[57] E. Deshayes, M. Roumiguie, C. Thibault, P. Beuzeboc, F. Cachin, C. Hen- nequin, D. Huglo, F. Rozet, D. Kassab-Chahmi, X. Rebillard, and N. Houédé. Radium 223 dichloride for prostate cancer treatment. Drug Des. Dev. Ther., 11:2643 – 2651, 2007.

[58] J. M. O’Sullivan. Radium-223 from bench to bedside – future directions for targeted alpha therapy. Eur. Oncol. Haematol., 14(Suppl. 2):3 – 9, 2018.

[59] M. D. Wissing, F. W. B. van Leeuwen, G. van der Pluijm, and H. Gelderblom. Radium-223 chloride: Extending life in prostate cancer patients by treating bone metastases. Clin. Cancer Res., 19(21):5822 –5827, 2013.

[60] C. Müller, N. P. van der Meulen, M. Benešová, and R. Schibli. Therapeutic radiometals beyond 177Lu and 90Y: Production and application of promising α-particle, β−-particle, and Auger electron emitters. J. Nucl. Med., 58(9):91S – 96S, 2017.

[61] M. R. McDevitt, D. Ma, L. T. Lai, J. Simon, P. Borchardt, R. K. Frank, K. Wu, V. Pellegrini, M. J. Curcio, M. Miederer, N. H. Bander, and D. A. Scheinberg. Tumor therapy with targeted atomic nanogenerators. Science, 294(5546):1537 – 1540, 2001.

[62] M. Miederer, G. Henriksen, A. Alke, I. Mossbrugger, L. Quintanilla- Martinez, R. Senekowitsch-Schmidtke, and M. Essler. Preclinical evaluation of the α-particle generator nuclide 225Ac for somatostatin receptor radio-

75 therapy of neuroendocrine tumors. Clin. Cancer Res., 14(11):3555 – 3561, 2008.

[63] P. E. Borchardt, R. R. Yuan, M. Miederer, M. R. McDevitt, and D. A. Scheinberg. Targeted actinium-225 in Vivo generators for therapy of ovarian cancer. Cancer Res., 63:5084 – 5090, 2003.

[64] C. Kratochwil, F. Bruchertseifer, F. L. Giesel, M. Weis, F. A. Verburg, F. Mottaghy, K. Kopka, C. Apostolidis, U. Haberkorn, and A. Morgenstern. 225Ac-PSMA-617 for PSMA-targeted α-radiation therapy of metastatic castration-resistant prostate cancer. J. Nucl. Med., 57(12):1941 – 1944, 2016.

[65] M. Miederer, D. A. Scheinberg, and M. R. McDevitt. Realizing the potential of the actinium-225 radionuclide generator in targeted alpha-particle therapy applications. Adv. Drug Deliv. Rev., 60(12):1371 – 1382, 2008.

[66] C. Kratochwil, F. L. Giesel, F. Bruchertseifer, W. Mier, C. Apostolidis, R. Boll, K. Murphy, U. Haberkorn, and A. Morgenstern. 213Bi-DOTATOC receptor-targeted alpha-radionuclide therapy induces remission in neuroendocrine tumours refractory to beta radiation: a ﬁrst-in-human experience. Eur. J. Nucl. Med. Mol. Imaging, 41:2106 – 2119, 2014.

[67] T. L. Rosenblat, D. A. Mulford M. R. McDevitt, N. Pandit-Taskar, C. R. Divgi, K. S. Panageas, M. L. Heaney, S. Chanel, A. Morgenstern, G. Sgouros, S. M. Larson, D. A. Scheinberg, and J. G. Jurcic. Sequential cytarabine and alpha-particle immunotherapy with bismuth-213-lintuzumab (HuM195) for acute myeloid leukemia. Clin. Cancer Res., 16(21):5303 – 5311, 2010.

[68] NuDat 2.7 database, https://www.nndc.bnl.gov/nudat2/.

[69] R. Gallicchio, S. Giacomobono, A. Nardelli, T. Pellegrino, V. Simeon, D. Gattozzi, F. Maddalena, P. Mainenti, and G. Storto. Palliative treat-

76 ment of bone metastases with samarium-153 EDTMP at onset of pain. J. Bone Miner. Metab., 32:434 – 440, 2014.

[70] A. Fakhari, A. R. Jalilian, H. Yousefnia, F. Johari-Daha, M. Mazidi, and A. Khalaj. Development of 166Ho-pamidronate for bone pain palliation therapy. J. Radioanal. Nucl. Chem., 303:743 – 750, 2015.

[71] M. Alavi, S. Omidvari, A. Mehdizadeh, A. R. Jalilian, and A. Bahrami-Samani. Metastatic bone pain palliation using 177Lu- ethylenediaminetetramethylene phosphonic acid. World J. Nucl. Med., 14(2):109 – 115, 2015.

[72] A. S. Shinto, D. Shibu, K. K. Kamaleshwaran, T. Das, S. Chakraborty, S. Banerjee, P. Thirumalaisamy, P. Das, and G. Veersekar. 177Lu-EDTMP for treatment of bone pain in patients with disseminated skeletal metastases. J. Nucl. Med. Tech., 42(1):55 – 61, 2014.

[73] S. J. Knox. Yttrium-90-labeled anti-CD20 monoclonal antibody therapy of recurrent B-cell lymphoma. Clin. Cancer Res., 2:457 – 470, 1996.

[74] R. D. Teo, J. Termini, and H. B. Gray. Lanthanides: Applications in cancer diagnosis and therapy. J. Med. Chem., 59(13):6012 – 6024, 2016.

[75] L. Emmett, K. Willowson, J. Violet, J. Shin, A. Blanksby, and J. Lee. Lutetium-177-PSMA radionuclide therapy for men with prostate cancer: a review of the current literature and discussion of practical aspects of therapy. J. Med. Radiat. Sci., 64:52 – 60, 2017.

[76] F. Rösch. Radiolanthanides in endoradiotherapy: an overview. Radiochim. Acta, 95:303 – 311, 2007.

[77] C. Müller, J. Reber, S. Haller, H. Dorrer, U. Köster, K. Johnston, K. Zher-

77 nosekov, A. Türler, and R. Schibli. Folate receptor targeted alpha-therapy using terbium-149. Pharmaceuticals, 7:353 – 365, 2014.

[78] C. Müller, C. Vermeulen, U. Köster, K. Johnston, A. Türler, R. Schibli, , and N. P. van der Meulen. Alpha-PET with terbium-149: evidence and perspectives for radiotheragnostics. EJNMMI Radiopharm. Chem., 1:5 – 9, 2016.

[79] J. Grünberg, D. Lindenblatt, H. Dorrer, S. Cohrs, K. Zhernosekov, U. Köster, A. Türler, E. Fischer, and R. Schibli. Anti-L1CAM radioimmunotherapy is more eﬀective with the radiolanthanide terbium-161 compared to lutetium- 177 in an ovarian cancer model. Eur. J. Nucl. Med. Mol. Imaging, 41:1907 – 1915, 2014.

[80] C. Müller, J. Reber, S. Haller, H. Dorrer, P. Bernhardt, K. Zhernosekov, A. Türler, and R. Schibli. Direct in vitro and in vivo comparison of 161Tb and 177Lu using a tumour-targeting folate conjugate. Eur. J. Nucl. Med. Mol. Imaging, 41:476 – 485, 2014.

[81] S. C. Srivastava. Paving the way to personalized medicine: Production of some promising theragnostic radionuclides at Brookhaven National Labora- tory. Semin. Nucl. Med., 42:151 – 163, 2012.

[82] R. A. Werner, C. Bluemel, M. S. Allen-Auerbach, T. Higuchi, and K. Her- rmann. 68Gallium- and 90Yttrium-/177Lutetium: "theranostic twins" for diagnosis and treatment of NETs. Ann. Nucl. Med., 29:1 – 7, 2015.

[83] L. F. Mausner and S. C. Srivastava. Selection of radionuclides for radioimmunotherapy. Med. Phys., 20(2):503 – 509, 1993.

[84] H. Herzog, F. Rösch, G. Stöcklin, C. Lueders, S. M. Qaim, and L. E. Feinen- degen. Measurement of pharmacokinetics of yttrium-86 radiopharmaceuti-

78 cals with PET and radiation dose calculation of analogous yttrium-90 radio- therapeutics. J. Nucl. Med., 34(12):2222 – 2226, 1993.

[85] S. Srivastava and E. Dadachova. Recent advances in radionuclide therapy. Semin. Nucl. Med., XXXI(4):330 – 341, 2001.

[86] C. Champion, M. A. Quinto, C. Morgat, P. Zanotti-Fregonara, and E. Hindié. Comparison between three promising β-emitting radionuclides, 67Cu, 47Sc and 161Tb, with emphasis on doses delivered to minimal residual disease. Theranostics, 6(10):1611 – 1618, 2016.

[87] C. Müller, M. Bunka, S. Haller, U. Köster, V. Groehn, P. Bernhardt, N. van der Meulen, A. Türler, and R. Schibli. Promising prospects for 44Sc-/47Sc-based theragnostics: Application of 47Sc for radionuclide tumor therapy in mice. J. Nucl. Med., 55(10):1658 – 1664, 2014.

[88] S. Lehenberger, C. Barkhausen, S. Cohrs, E. Fischer, J. Grünberg, U. Köster A. Hohn and, R. Schibli, A. Türler, and K. Zhernosekov. The low-energy β− and electron emitter 161Tb as an alternative to 177Lu for targeted radionuclide therapy. Nucl. Med. Biol., 38:917 – 924, 2011.

[89] C. Müller, K. A. Domnanich, C. A. Umbricht, and N. P. van der Meulen. Scandium and terbium radionuclides for radiotheranostics: current state of development towards clinical application. Br. J. Radiol., 91:1 – 13, 2018.

[90] P. Auger. Sur l’eﬀet photoélectrique composé. J. Phys. Radium, 6(6):205 – 208, 1925.

[91] https://commons.wikimedia.org/wiki/category:electron_capture#/media /ﬁle:elektroneneinfang_(2_phasen).png.

79 [92] A. I. Kassis. The amazing world of Auger electrons. Int. J. Radiat. Biol., 80(11 – 12):789 – 803, 2004.

125 [93] T. A. Carlson and R. M. White. Formation of fragment ions from CH3Te 125 125 125 and C2H5Te following the nuclear decays of CH3I and C2H5I . J. Chem. Phys., 38:2930 – 2934, 1963.

[94] A. I. Kassis, K. S. R. Sastry, and S. J. Adelstein. Kinetics of uptake, retention, and radiotoxicity of 125IUdR in mammalian cells: Implications of localized energy deposition by Auger processes. Radiat. Res., 109(1):78 – 89, 1987.

[95] G. M. Makrigiorgos, A. I. Kassis, J. Baranowska-Kortylewicz, K. D. McElvany, M. J. Welch, K. S. R. Sastry, and S. J. Adelstein. Ra- diotoxicity of 5-[123i]Iodo-20-Deoxyuridine in V79 cells: A comparison with 5-[125i]Iodo-20-Deoxyuridine. Radiat. Res., 118(3):532 – 544, 1989.

[96] A. I. Kassis, S. J. Adelstein, C. Haydock, K. S. R. Sastry, K. D. McElvany, and M. J. Welch. Lethality of Auger electrons from the decay of Bromine-77 in the DNA of mammalian cells. Radiat. Res., 90(2):362 – 373, 1982.

[97] E. R. DeSombre, P. V. Harper, A. Hughes, R. C. Mease, S. J. Gatley, O. T. DeJesus, and J. L. Schwartz. Bromine-80m radiotoxicity and the potential for estrogen receptor-directed therapy with Auger electrons. Cancer Res., 48:5805 – 5809, 1988.

[98] R. W. Howell. Auger processes in the 21st century. Int. J. Radiat. Biol., 84(12):959 – 975, 2008.

[99] L. Y. Xue, N. J. Butler, G. M. Makrigiorgos, S. J. Adelstein, and A. I. Kassis. Bystander eﬀect produced by radiolabeled tumor cells in vivo. Proc. Natl. Acad. Sci., 99(21):13765 – 13770, 2002.

80 [100] O. Mamlouk, P. Balagurumoorthy, K. Wang, S. J. Adelstein, and A. I. Kas- sis. Bystander eﬀect in tumor cells produced by iodine-125 labeled human lymphocytes. Int. J. Radiat. Biol., 88(12):1019 – 1027, 2012.

[101] C. A. Boswell and M. W. Brechbiel. Auger electrons: Lethal, low energy, and coming soon to a tumor cell nucleus near you. J. Nucl. Med., 46:1946 – 1947, 2005.

[102] S. J. Adelstein and A. I. Kassis. Strand breaks in plasmid Dna following positional changes of Auger-electron-emitting radionuclides. Acta Oncol., 35(7):797 – 801, 1996.

[103] B. Cornelissen and K. A. Vallis. Targeting the nucleus: An overview of Auger-electron radionuclide therapy. Curr. Drug Discov. Technol., 7, 2010.

[104] D. V. Rao, P. N. Goodwin, and F. L. Khalil. 165Er: An ”ideal” radionuclide for imaging with pressurized multiwire proportional gamma cameras. J. Nucl. Med., 15(11):1008 – 1010, 1974.

[105] A. Parr, R. M. Beihn, and M. Jay. Radiolabeling of an enteric coated tablet by (n,γ) radioactivation of erbium-170 for scintigraphic imaging. Int. J. Pharm., 32:251 – 256, 1986.

[106] R. Chakravarty, S. Chakraborty, V. Chirayil, and A. Dash. Reactor production and electrochemical puriﬁcation of 169Er: A potential step forward for its utilization in in vivo therapeutic applications. Nucl. Med. Biol., 41:163 – 170, 2014.

[107] F. Tárkányi, A. Hermanne, S. Takács, F. Ditrói, B. Király, S. F. Kovalev, and A. V. Ignatyuk. Experimental study of the 165Ho(p,n) nuclear reaction for production of the therapeutic radioisotope 165Er. Nucl. Instrum. Methods. Phys. Res. B, 266:3346 – 3352, 2008.

81 [108] F. Tárkányi, A. Hermanne, S. Takács, F. Ditrói, B. Király, S. F. Kovalev, and A. V. Ignatyuk. Experimental study of the 165Ho(d,2n) and 165Ho(d,p) nuclear reactions up to 20 mev for production of the therapeutic radioisotopes 165Er and 166gHo. Nucl. Instrum. Methods. Phys. Res. B, 266:3529 – 3534, 2008.

[109] F. Tárkányi, S. Takács, A. Hermanne, F. Ditrói, B. Király, M. Baba, T. Oht- suki, S. F. Kovalev, and A. V. Ignatyuk. Investigation of production of the therapeutic radioisotope 165Er by proton induced reactions on erbium in comparison with other production routes. Appl. Radiat. Isot., 67:243 – 247, 2009.

[110] F. Tárkányi, A. Hermanne, B. Király, S. Takács, F. Ditrói, M. Baba, T. Oht- suki, S. F. Kovalev, and A. V. Ignatyuk. Study of activation cross-sections of deuteron induced reactions on erbium: Production of radioisotopes for practical applications. Nucl. Instrum. Methods. Phys. Res. B, 259:829 – 835, 2007.

[111] F. Tárkányi, A. Hermanne, S. Takács, B. Király, I. Spahn, and A. V. Ig- natyuk. Experimental study of the excitation functions of proton induced nuclear reactions on 167Er for production of medically relevant 167Tm. Appl. Radiat. Isot., 68:250 – 255, 2010.

[112] A. Hermanne, R. Adam-Rebeles, F. Tarkanyi, S. Takacs, J. Csikai, M. P. Takacs, and A. Ignatyuk. Deuteron induced reactions on Ho and La: Ex- perimental excitation functions and comparison with code results. Nucl. Instrum. Methods. Phys. Res. B, 311:102 – 111, 2013.

[113] M. Sadeghi, M. Enferadi, and C. Tenreiro. Nuclear model calculations on the production of Auger emitter 165Er for targeted radionuclide therapy. J.

82 Mod. Phys., 1:217 – 225, 2010.

[114] W. T. Leland. The isotopic constitution of scandium, gadolinium, dyspro- sium, holmium, erbium, and ytterbium. Phys. Rev., 77(5):634 – 640, 1950.

[115] N. Zandi, M. Sadeghi, and H. Afarideh. Evaluation of the cyclotron production of 165Er by diﬀerent reactions. J. Radioanal. Nucl. Chem., 295:923 – 928, 2013.

[116] Large Research Facilities. https://www.psi.ch/science/large-research- facilities.

[117] P. Janoš. Analytical separations of lanthanides and actinides by capillary electrophoresis. Electrophoresis, 24:1982 – 1992, 2003.

[118] X.-Z. Li. A supramolecular lanthanide separation approach based on mul- tivalent cooperative enhancement of metal ion selectivity. Nat. Commun., 9:547 – 556, 2018.

[119] F. Xie, T. A. Zhang, D. Dreisinger, and F. Doyle. A critical review on solvent extraction of rare earths from aqueous solutions. Miner. Eng., 56:10 – 28, 2014.

[120] E. M. I. Zubiani, C. Cristiani, G. Dotelli, and P. G. Stampino. Solid liquid extraction of rare earths from aqueous solutions: A review. Procedia Environ. Sci. Eng. Manag., 2(3):231 – 237, 2015.

[121] N. van der Meulen. Ion exchange behaviour of 42 selected elements on AG MP-50 cation exchange resin in nitric acid and citric acid mixtures. Master’s thesis, University of Stellenbosch, 2003.

[122] E. P. Horwitz, D. R. McAlister, and M. L. Dietz. Extraction chromatography versus solvent extraction: How similar are they? Sep. Sci. Technol., 41:2163

83 – 2182, 2006.

[123] D. McAlister and P. Horwitz. Characterization of extraction of chromatographic materials containing bis(2-ethyl-1-hexyl)phosphoric acid, 2- ethyl-1-hexyl(2-ethyl-1-hexyl)phosphonic acid, and bis(2,4,4-trimethyl-1- pentyl)phosphinic acid. Solvent Extr. Ion Exch., 25(6):757 – 769, 2007.

[124] F. Monroy-Guzman and E. J. Salinas. Separation of micro-macrocomponent systems: 149Pm–Nd, 161Tb–Gd, 166Ho–Dy and 177Lu–Yb by extraction chromatography. J. Mex. Chem. Soc., 59(2):143 – 150, 2015.

[125] Andrey G. Kazakov, R. A. Aliev, A. Yu. Bodrov, A. B. Priselkova, and S. N. Kalmykov. Separation of radioisotopes of terbium from a europium target irradiated by 27 mev α-particles. Radiochim. Acta, 106(2):135 – 140, 2018.

[126] E. P. Horwitz, D. R. McAlister, A. H. Bond, R. E. Barrans, and J. M. Williamson. A process for the separation of 177lu from neutron irradiated 176yb targets. Appl. Radiat. Isot., 63:23 – 36, 2005.

[127] E. R. Tompkins and S. W. Mayer. Ion exchange as a separations method. iii. equilibrium studies of the reactions of rare earth complexes with synthetic ion exchange resins. J. Am. Chem. Soc., 69(11):2859 – 2865, 1947.

[128] J. Korkisch. Modern Methods for the Separation of Rarer Metal Ions. Else- vier, 1 edition, 1969.

84 A Results of inductively coupled plasma optical emission spectroscopy measurements

Table A.1: Results of ICP–OES measurements carried out to determine the retention capacities of lanthanide of interest on LN1 resin. Unit for all values is ppm.

Ho Er Tm 1 2 3 1 2 3 1 2 3

f1 0.623 0.052 0.065 2.909 0.621 0.053 0.537 0.427 0.065

f2 0.069 0.024 0.036 0.265 0.074 0.028 0.060 0.046 0.018

f3 0.026 0.010 0.014 0.079 0.026 0.016 0.022 0.032 0.010

f4 0.012 0.004 0.008 0.022 0.011 0.006 0.010 0.013 0.005

f5 0.006 0.003 0.005 0.005 0.002 0.003 0.003 0.007 0.002

f6 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000

e1 0.724 0.582 0.952 0.551 0.720 0.624 0.791 0.676 0.427

e2 1.372 1.304 1.316 1.149 1.155 1.074 1.108 1.014 0.585

e3 0.750 0.726 0.778 0.414 0.449 0.398 0.196 0.285 0.176

e4 0.001 0.009 0.011 0.003 -0.015 0.000 -0.011 0.005 -0.007

e5 -0.017 -0.012 -0.010 -0.020 -0.026 -0.010 -0.015 -0.012 -0.010

e6 -0.019 -0.014 -0.014 -0.023 -0.032 -0.013 -0.016 -0.015 -0.012

e’1 1.041 0.653 0.533 0.709 0.665 0.832 0.416 0.741 0.447

e’2 1.330 0.920 1.003 0.963 0.955 1.188 0.957 1.331 0.531

e’3 1.719 1.211 1.301 1.186 1.406 1.483 1.189 1.517 0.715

e’4 1.889 1.354 1.527 1.332 1.610 1.546 1.344 1.437 1.170

e’5 1.752 1.627 1.469 1.521 1.634 1.579 1.319 1.552 1.194

e’6 1.855 1.491 1.580 1.486 1.531 1.923 1.287 1.307 1.218 ref 48.87 48.91 48.54 54.35 49.37 52.63 49.23 49.36 54.69

85 Table A.2: Results of ICP–OES measurements carried out to determine the retention capacities of lanthanide of interest on LN2 resin. Unit for all values is ppm.

Ho Er Tm 1 2 3 1 2 3 1 2 3

f1 0.794 0.045 0.030 0.048 0.053 0.045 0.053 0.063 0.127

f2 0.096 0.021 0.017 0.018 0.023 0.024 0.020 0.026 0.017

f3 0.029 0.011 0.007 0.009 0.029 0.014 0.016 0.013 0.008

f4 0.009 0.004 0.002 0.004 0.007 0.006 0.008 0.007 0.019

f5 0.004 0.002 0.001 0.002 0.001 0.004 0.001 0.000 0.001

f6 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.001 0.000

e1 1.860 1.889 1.751 4.735 1.807 1.980 2.011 1.889 2.170

e2 6.200 6.102 5.850 15.16 6.056 6.545 6.985 6.223 6.765

e3 8.861 9.383 8.283 16.61 8.525 8.763 7.156 6.836 6.657

e4 1.120 0.996 0.913 0.492 0.514 0.463 0.205 0.141 0.212

e5 0.093 0.101 0.198 0.022 0.029 0.047 0.004 -0.004 0.005

e6 -0.021 -0.008 -0.006 -0.011 -0.012 -0.008 -0.014 -0.015 -0.009

e’1 0.049 0.085 0.054 0.260 0.059 0.103 0.056 0.060 0.044

e’2 0.029 0.054 0.029 0.249 0.047 0.033 0.077 0.098 0.104

e’3 0.118 0.120 0.106 0.791 0.213 0.323 0.889 0.668 0.980

e’4 2.037 2.078 1.893 2.346 2.135 2.294 2.476 2.159 2.598

e’5 2.143 2.196 2.017 2.215 2.230 2.416 2.489 2.191 2.548

e’6 2.294 2.226 2.101 2.472 2.246 2.458 2.472 2.189 2.600 ref 52.08 49.26 50.08 51.92 49.38 51.61 57.60 48.26 59.16

86 Table A.3: Results of ICP–OES measurements carried out to determine the retention capacities of lanthanide of interest on LN3 resin. Unit for all values is ppm.

Ho Er Tm 1 2 3 1 2 3 1 2 3

f1 0.497 0.081 0.029 0.445 0.146 0.046 0.843 0.044 0.013

f2 0.062 0.027 0.010 0.058 0.038 0.025 0.084 0.022 0.006

f3 0.025 0.009 0.009 0.023 0.013 0.013 0.046 0.012 0.002

f4 0.011 0.008 0.000 0.009 0.004 0.007 0.008 0.134 0.002

f5 0.005 0.000 0.000 0.003 0.000 0.005 0.000 0.006 0.000

f6 0.000 0.006 0.002 0.000 0.000 0.000 0.005 0.000 0.000

e1 4.766 1.955 1.811 4.551 1.962 2.091 2.082 1.853 2.002

e2 9.576 6.384 5.781 9.945 6.251 6.946 6.964 6.081 6.384

e3 14.86 9.559 8.700 13.53 9.375 10.44 10.21 9.211 9.663

e4 20.73 19.70 17.60 19.24 19.04 20.46 20.35 18.71 18.76

e5 14.19 19.11 16.32 14.60 16.52 17.63 14.06 11.86 13.03

e6 0.942 0.679 0.420 0.562 0.354 0.161 0.102 0.049 0.168

e’1 0.169 0.337 0.193 0.053 0.321 0.349 0.305 0.240 0.286

e’2 0.955 0.139 0.157 0.666 0.119 0.171 0.106 0.104 0.227

e’3 0.908 0.124 0.136 0.76 0.120 0.155 0.104 0.083 0.214

e’4 1.295 0.129 0.133 1.119 0.146 0.196 0.194 0.131 0.309

e’5 1.608 0.196 0.290 1.417 0.420 0.527 0.842 0.851 0.951

e’6 2.311 2.241 2.051 1.943 2.259 2.478 2.444 2.116 2.395 ref 50.20 51.59 52.08 47.39 49.73 54.03 55.42 50.21 53.38