ISOTOPE HARVESTING OF AQUEOUS PHASE IONS FROM HEAVY - ION FRAGMENTATION FACILITIES FOR THE PRODUCTION OF A 47 Ca/ 47 Sc GENERATOR By Emily Paige Abel A DISSERTATION Submitted to Michigan State University i n partial fulfillment of the requirem ents for the degree of Chemistry Doctor of Philosophy 2020 ABSTRACT ISOTOPE HARVESTING OF AQUEOUS PHASE IONS FROM HEAVY - ION FRAGMENTATION FACILITIES FOR THE PRODUCTION OF A 47 Ca/ 47 Sc GENERATOR By Emily Paige Abel Targeted internal radio therapy is a promising new treatment method for metastatic cancer s . Several scandium isotopes, 43 , 44 , 47 Sc, could serve as paired diagnostic and therapeutic isotopes for these diseases . While ideal production routes for diagnostic 43,44 Sc have been found using small medical cyclotrons, research to find a sustainable production method for the therapeutic isotope 47 Sc is ongoing . At the National Superconducting Cyclotron Laboratory (NSCL) and in the future at the Facility for Rare Isotope Be a ms (FRIB) , a supply of 47 Ca , the parent of 47 Sc, can be produced on a regular b asis when a 48 Ca primary beam is used. While the main function of th e accelerated primary beam is to produce a user - specified secondary radioactive beam for nuclear physics e xp eriments, >90% of th e primary beam goes unreacted and is collecte d in beam stops. When a beam blocker with a water interior is used, t he 48 Ca primary beam produces 47 Ca as the most abundant fragment resulting from nuclear reactions in the bea m blocker . Thi s method of production allows for th e use of unreacted, accelerated beams for isotope production. In the future at FRIB, production of 47 Ca through isotope harvesting is predicted to reach the TBq/day level when a 48 Ca primary b eam is in use. This supply of 47 Ca will facility the production of 47 Ca/ 47 S c generators for research quantities of 47 Sc. Isotope harvesting methods for the generation of 47 Sc, among other interesting products available through this production method, ha ve been explored through several experiments at the NSCL and the Cyclotron La b at the University of Wisconsin - Madison. An isotope harvesting system has been developed to include a flowing - water target, components to condition and monitor the water, and c om po nents to collect radionuclidic products of interest. The durability of this system to irradiation conditions has been tested using a low intensity 40 Ca irradiation and a high intensity proton irradiation. Through 48 Ca irradiations at the NSCL, m ethods to c ollect and purify 47 Ca have been tested and optimized . Using these methods, samples of 47 Sc with high radionuclidic purity have been generated and used to radiolabel DTPA - TOC, a biologically active molecule used to target neuroendocrine tumors. This wo rk h as verified the feasibility of using isotope harvesti ng to generate 47 Sc for nuclear medicine applications. iii ACKNOWLEDGEMENTS As I was thinking of the list of people I would like to acknowledge for helping me along this path, I felt grateful for al l ds and family and colleagues. I would like to start by thanking the entire isotope harvesting group. While our group has grown from four members at the start to thirteen members currently, the support and camar ad er tant. Thank you all for the time and given making the experiments in this work a success, listening to my practice talks, reading drafts of papers, and making out group activities so much fun. I wo ul d also like to thank a few people in parti cular. Thank you, Hannah, for all of the hours we spent together in lab during our first few years of grad school. We accidently sprayed each other with water from the harvesting system more times than I can co un t, but we also learned a lot together. I wou ld also like to give a huge thanks to Kathi for all the scandium chemistry expertise she brought to our group and all the time and ideas she contributed to the experiments presented in this work. Thank you, Kat hi , for spending so much time in lab with me! I would like to give a shout out to Hannah, Kathi, and Sam for being awesome office mates. I love d our conversations about work and life in general. I also hope Colton and Wes know how grateful I am for their d ed ic ation to building an isotope harvesting sy stem and keeping it running for experiments no matter what problems they faced. Thank you for all the long hours you both contributed and the safety hoops you jumped through for all of our experiments! I wou ld a lso like to acknowledge the faculty suppor t I received. Thank you to my committee for their constructive comments at each of our meetings. iv support of my undergraduate advisors, Dr. Feigl and Dr. Barstis. They encouraged my int er es t in research and my goal of going to grad school. Thank you, Dr. Feigl, for suggesting I explore nuclear chemistry! much your support and advice has meant to me. Your guidan ce h as always been patient , kind , and encouraging, especially when I struggled to have confidence in myself . Thank you for asked for a better advisor! I feel lucky to h ave such a supportive family. My p arents have always supported my education and my dream to be a chemist from the age of 12. Thank you, Mom and Dad, for encouraging me when things got tough and I started to doubt that dream . I would also like to thank my s isters for being my lifelong frien He always had a yummy dinner and a warm hug for me when I got home from a 15 - g, T ag, who was my buddy through grad school. For all of these things and so many more, I would like to thank everyone who has helped me reach this goal. Now, I hope to use all of the others with my passion f or chemistry. v TABLE OF CONTENTS LIST OF TABLES ................................ ................................ ................................ ................................ ix LIST OF FIGURES ................................ ................................ ................................ ............................. xii Chapter 1 : Introduction ................................ ................................ ................................ ................. 1 Chapter 2 : Development of Isotope Harvesting System ................................ ............................. 16 2.1 Introduction ................................ ................................ ................................ ........................ 16 2.2 Target Design ................................ ................................ ................................ ...................... 22 2.2.1 Generation 1: Low Intensity 40 Ca Irradiation ................................ ............................... 22 2.2.2 Generation 2: 1 H and 48 Ca Irradiations ................................ ................................ ........ 24 2.3 Water System Design ................................ ................................ ................................ .......... 26 2.3.1 Generation 1: Low Intensity 40 Ca Irradiation ................................ ............................... 26 2 .3.1.1 Water System ................................ ................................ ................................ ...................... 26 2.3.1.2 Conclusions about the Generation 1 Water System ................................ ........................... 30 2.3.2 G eneration 2 and 3: 1 H and Low Intensity 48 Ca Irradiation ................................ ......... 31 2.3.3 Generation 4: High Intensity 48 Ca Irradiations ................................ ............................. 39 2.4 Conclusion ................................ ................................ ................................ ........................... 43 Chapter 3 : Low Intensity 40 Ca Beam Ti64 Durability Test ................................ ............................ 45 3.1 Introduction ................................ ................................ ................................ ........................ 45 3.2 Materials and Methods ................................ ................................ ................................ ....... 48 3.2.1 Isotope Harvesting System Irradiation ................................ ................................ ........ 48 3.2.2 Radionuclide Quantification and Compa rison to Production Estimates ..................... 51 3.2.3 Radiolysis Measurement ................................ ................................ .............................. 54 3.2.4 Corrosion Assessment ................................ ................................ ................................ .. 56 3.2.4.1 Ion Exchange Resin Processing ................................ ................................ .............. 56 3.2.4.2 Surface Assessment ................................ ................................ ............................... 57 3.2.4.3 Corrosion Rate Estimation ................................ ................................ ..................... 58 3.3 Results and Discussion ................................ ................................ ................................ ........ 60 3.3.1 Isotope Harvesting System Irradiation ................................ ................................ ........ 60 3.3.2 Radionuclide Quantification and Comparison to Production Estimates ..................... 64 3.3.3 Radiolysis Measurement ................................ ................................ .............................. 69 3.3.4 Corrosion Assessment ................................ ................................ ................................ .. 72 3.3.4.1 Ion Exchange Resin Processing ................................ ................................ .............. 72 3.3.4.2 Surface Asse ssment ................................ ................................ ............................... 76 3.3.4.3 Corrosion Rate Estimation ................................ ................................ ..................... 77 3.4 General Discussion ................................ ................................ ................................ .............. 78 3.5 Conclusion ................................ ................................ ................................ ........................... 80 vi Chapter 4 : Low Intensity 48 Ca Irradiation for the Production, Collection, and Purification of 47 Ca ................................ ................................ ................................ ................................ ....................... 82 4.1 Introduction ................................ ................................ ................................ ........................ 83 4.2 Materials and Methods ................................ ................................ ................................ ....... 84 4.2.1 Materials ................................ ................................ ................................ ...................... 84 4.2.1.1 Reagents ................................ ................................ ................................ ................. 84 4.2.1.2 Extraction Chromatography and Ion Exchange Resins ................................ .......... 84 4.2.1.3 Column Construction ................................ ................................ ............................. 85 4.2.1.4 Instruments ................................ ................................ ................................ ............ 86 4.2.2 48 Ca Irradiation ................................ ................................ ................................ ............. 87 4.2.3 Production of 47 Ca ................................ ................................ ................................ ........ 89 4.2.4 Collection and Sample Processing ................................ ................................ ............... 91 4.2.5 Purification of 47 Ca ................................ ................................ ................................ ....... 92 4.2.5.1 Separation Method 1: DGA resin with HCl and HNO 3 ................................ ........... 92 4.2.5.2 Separation Method 2: AG MP - 50 with HCl ................................ ............................ 92 4.2.5.3 Se paration Method 3: AG MP - 50 with Methanolic HCl ................................ ........ 93 4.2.5.4 Separation Yield and Radionuclidic Purity ................................ ............................. 94 4.2.6 Stable Elemental Analysis ................................ ................................ ............................ 95 4 .3 Results and Disc ussion ................................ ................................ ................................ ........ 96 4.3.1 Production of 47 Ca ................................ ................................ ................................ ........ 96 4.3.2 Collection and Sample Processing ................................ ................................ ............... 99 4 .3.3 Purification of 47 Ca ................................ ................................ ................................ ..... 101 4.3.3.1 Separation Methods ................................ ................................ ............................ 101 4.3.3.2 Separation Yield and Radionuclidic Purity ................................ ........................... 107 4.3.4 Comparison of Separation Methods ................................ ................................ .......... 108 4.3.5 Stable Eleme ntal Analysis ................................ ................................ .......................... 110 4.4 Conclusion ................................ ................................ ................................ ......................... 113 Chapter 5 : Higher Intensity 1 H Beam Target Durability Test ................................ ..................... 114 5.1 Introduction ................................ ................................ ................................ ...................... 114 5.2 Experimental Methods ................................ ................................ ................................ ..... 117 5.2.1 Experiment Design ................................ ................................ ................................ ..... 117 5.2.2 Quantification of Radionuclides ................................ ................................ ................ 120 5.2.2.1 Radiation Measurements for the Quantification of Radionuclides ..................... 120 5.2.2.2 Estimated Production of Radionuclides ................................ ............................... 123 5.2.3 Estimating Activity of 48 V and 51 Cr in the System Water from Nuclear Recoil .......... 123 5.2.4 Degradation of the Target ................................ ................................ ......................... 126 5.3 Results and Discussion ................................ ................................ ................................ ...... 129 5.3.1 Quantification of Radionuclides Produced in Water Target ................................ ...... 129 5.3.2 Degradation of the Target ................................ ................................ ......................... 131 5.3.2.1 Estimating the Corrosion Rate ................................ ................................ ............. 131 5.3.2.2 Extrapolating Corrosion Rate t o Predict Target Lifetimes ................................ ... 134 5.3.2.3 Validity and Limitations of Extrapolation ................................ ............................ 138 5.4 Conclusion ................................ ................................ ................................ ......................... 141 vii Chapter 6 : 48 Ca Beam Experiment 2: Proof of Concept for 47 Ca/ 47 Sc Generator and 47 Sc Radiolabeling with Isotope Harvested 47 Ca ................................ ................................ ................ 142 6.1 Introduction ................................ ................................ ................................ ...................... 143 6.2 Materials and Methods ................................ ................................ ................................ ..... 144 6 .2.1 Materials ................................ ................................ ................................ .................... 144 6.2.1.1 Chemicals and Resi ns ................................ ................................ ........................... 144 6.2.1.2 Instruments ................................ ................................ ................................ .......... 146 6.2.2 48 Ca Irradiation ................................ ................................ ................................ ........... 146 6.2.3 Produc tion of 47 Ca ................................ ................................ ................................ ...... 147 6.2.4 Collection of 47 Ca from Isotope Harvesting System ................................ .................. 149 6.2.5 Purification of 47 Ca ................................ ................................ ................................ ..... 151 6.2.6 Generation of 47 Sc ................................ ................................ ................................ ...... 154 6.2.7 Radiolabeling DTPA - TOC with 47 Sc ................................ ................................ ............. 157 6.2.8 Stable Elemental Analysis ................................ ................................ .......................... 161 6.3 Results and Discussion ................................ ................................ ................................ ...... 162 6.3.1 Production of Radionucli des with 48 Ca Beam ................................ ............................ 162 6.3.2 Collection of 47 Ca from Isotope Harvesting System ................................ .................. 164 6.3.3 Purification of 47 Ca ................................ ................................ ................................ ..... 167 6 .3.4 Generat ion of 47 Sc ................................ ................................ ................................ ...... 168 6.3.5 Radiolabeling DTPA - TOC with 47 Sc ................................ ................................ ............. 171 6.3.6 Stable Elemental Analysis ................................ ................................ .......................... 173 6.4 Conclusion ................................ ................................ ................................ ......................... 179 Chapter 7 : 48 Ca Beam Experiment 3: High Activity 47 Ca/ 47 Sc Generator and 47 Sc Radiolabeling with Isotope Harvested 47 Ca ................................ ................................ ................................ ....... 180 7.1 Introduction ................................ ................................ ................................ ...................... 180 7.2 Materials and Methods ................................ ................................ ................................ ..... 181 7.2.1 Materials ................................ ................................ ................................ .................... 181 7.2.2 48 Ca Irradiation ................................ ................................ ................................ ........... 181 7.2.3 Collection of 47 Ca from Isotope Harvesting System ................................ .................. 182 7.2.4 Purification of 47 Ca ................................ ................................ ................................ ..... 183 7.2.5 Generation of 47 Sc ................................ ................................ ................................ ...... 184 7.2.6 Radiolabeling DTPA - TOC with 47 Sc ................................ ................................ ............. 185 7.2.7 Stable Elemental Analysis ................................ ................................ .......................... 186 7.3 Results and Discussion ................................ ................................ ................................ ...... 187 7.3.1 Collection of 47 Ca from Isotope Harvesting System ................................ .................. 187 7.3.2 Purification of 47 Ca ................................ ................................ ................................ ..... 188 7.3.3 Generation of 47 Sc ................................ ................................ ................................ ...... 190 7.3.4 Radiolabeling DTPA - TOC with 47 Sc ................................ ................................ ............. 192 7.3.5 Stable Elemental Analysis ................................ ................................ .......................... 193 7.4 Conclusion ................................ ................................ ................................ ......................... 196 viii Chapter 8 : Measurement of the Three Most Intense Gamma Rays Following the Decay of 47 Ca ................................ ................................ ................................ ................................ ..................... 197 8.1 Introduction ................................ ................................ ................................ ...................... 197 8.2 Methods ................................ ................................ ................................ ............................ 199 8.2.1 Production of 47 Ca ................................ ................................ ................................ ...... 199 8.2.2 Purification of 47 Ca ................................ ................................ ................................ ..... 200 8.2.3 LSC Measurements ................................ ................................ ................................ .... 201 8.2.4 HPGe Gamma - Ray Spectroscopic Measurements ................................ ..................... 203 8.2.5 Error Budget ................................ ................................ ................................ ............... 205 8.3 Results and Discussion ................................ ................................ ................................ ...... 207 8. 4 Conclusion ................................ ................................ ................................ ......................... 216 Chapter 9 : General Discussion ................................ ................................ ................................ ... 217 9.1 Introduction ................................ ................................ ................................ ...................... 217 9.2 Production Rate of 47 Ca ................................ ................................ ................................ .... 217 9. 3 Separation Procedures for the Purification of 47 Ca ................................ .......................... 220 9.4 Stable Elemental Analysis ................................ ................................ ................................ . 223 9.5 Conclusion ................................ ................................ ................................ ......................... 224 APPENDICES ................................ ................................ ................................ ................................ 225 APPENDIX A: NUCLEAR DATA ................................ ................................ ................................ . 226 APPENDIX B: PREDICTED NUCLEAR RE ACTION PROBABIL ITIES ................................ .............. 229 APPENDIX C: STABLE ELEMENTAL ANALYSIS ................................ ................................ .......... 243 APPENDIX D: PRODUCTION RATE MEAUSREMENTS IN CHAPTER 5 ................................ ....... 245 REFERENCES ................................ ................................ ................................ ................................ 249 ix LIST OF TABLES Table 3.1: Quantification of and Production Rate Estimate for Radionuclides Pr oduced in Water Target ................................ ................................ ................................ ................................ ............ 65 Table 3.2: Activity Measured and Predi cted for Radionuclides Produced through Fragmentation Reactions. ................................ ................................ ................................ ................................ ...... 67 Table 3.3: Activity Measured and Predicted for Radionuclides Pro duced through Fusion Evaporation Reactions. ................................ ................................ ................................ ................. 68 Table 3.4: Avera ge Hydrogen Gas Production and Experimental G - value for Six Irra diation Periods ................................ ................................ ................................ ................................ .......... 71 Table 3.5: Quantification of Stable Elements Elute d from Ion Exchange Resins after Irradiation, Collection, and Elution ................................ ................................ ................................ .................. 73 Table 4.1: 47 Ca Activity Measured in Each Water Sample and C ation Exchange Resin Bed ......... 97 Table 4.2: Predicted and Measured Production Rates of 47 Ca in Isotope Harvesting Water Target with a 140 MeV/nucleon 48 Ca Beam ................................ ................................ ............................. 98 Table 4.3: Quantification of Radionuclides Collected on Cation Exchange Resins 1 - 3 .............. 101 Table 4.4: Example Replicate with Separation Method 1: DGA with 3 M HNO 3 /3M HCl .......... 104 Table 4.5: Example Replicate with Separation Method 2: AG MP - 50 with HCl Gradient .......... 105 Table 4.6: Example Replicate with Separation Method 3: AG MP - 50 with HCl/Methanol Gradient ................................ ................................ ................................ ................................ ..................... 106 Table 4.7: Sep aration Yield and Radionuclidic Purity of 47 Ca for Three Separation Methods ... 107 Table 4.8 : Stable Element Semi - Quantification ................................ ................................ .......... 111 Table 5.1: Summary of Experiments Performed with T i64 Target Material and Isotope Harvesting Water System ................................ ................................ ................................ ........... 116 Table 5.2: Estimated Production Rates for Radionuclides Produced in Water and Target Shell 122 Table 5.3: 13 N and 18 F Activities Produced in the Flowing - Water Target ................................ ... 129 Table 5.4: 48 V and 51 Cr Activity in the Water ................................ ................................ .............. 131 x Table 5.5: Scaled Co rrosion Rates for Different Beam Power Considerations ........................... 135 Table 6.1: Radiolabeling Conditions ................................ ................................ ........................... 159 Table 6.2: Production Rates of Radionuclides in Flowing - Water Target with 140 MeV/nucleon 48 Ca Beam ................................ ................................ ................................ ................................ .... 163 Table 6.3: Generator 1 to 6 Results ................................ ................................ ............................ 169 Table 6.4 : Radiolabeling Results ................................ ................................ ................................ . 171 Table 6.5 : ICP - OES Analysis of Eluate from Cation Exchange Resin Bed 1 ................................ . 176 Table 6.6 : ICP - OES Analysis of Eluate from Cation Exchange Resin Bed 2 ................................ . 177 Table 6.7 : ICP - OES Analysis of Samples from the Purification of 47 Ca and 47 Sc ......................... 178 Table 7.1: Radiolabeling Conditions ................................ ................................ ........................... 186 Table 7.2: Stable K + /Ca 2+ Test Separations ................................ ................................ ................. 187 Table 7.3: G enerator 1 - 3 Results ................................ ................................ ................................ 190 Table 7.4: Radiolabeling Results ................................ ................................ ................................ . 192 Table 7.5: Stable Element Results from Generators 1 - 4 and AG MP - 50 # 2 Separation ............. 195 Table 8.1: Half - Lives of 47 Ca and 47 Sc ................................ ................................ .......................... 208 Table 8.2: Error Budget for Half - lives Found from 159.4 keV Gamma - Ray Peak ....................... 209 Table 8.3: Error Budget for Three 47 Ca Branching Ratio s ................................ ........................... 213 Table 8.4: Minor Gamma - Ray Branching Ratio Values ................................ ............................... 214 Table 9.1: Comparison of Measured to Calculated 47 Ca Activities ................................ ............. 218 Table A1: Nuclear Data Used for Identifica tion, Quantification, and Localization of Radionuclides Produced with a 40 Ca Beam in a Flowing - Water Target. [14,15,28,50 56] ............................... 226 Table A2 : Nuclear Data Used to Quantify Radionuclides Produced By a 47 Ca Beam in a Flowing - Water Target 6,7,11,21,22,30 - 34 ................................ ................................ ................................ ........... 227 xi Table A3: Nuclear Data and Geometry Correction Factor for Production Rate Measurement . 228 Table B1 : Cross Section Data from PACE4 for the Production of 43 Sc through the 16 O + 40 Ca Fusion Evaporation Reaction. ................................ ................................ ................................ ..... 231 Table B2 : Cross Sect ion D ata from PACE4 for the Production of 44 Sc through the 16 O + 40 Ca Fusion Evaporation Reaction ................................ ................................ ................................ ...... 232 Table B3: Cross Section Data from PA CE4 for the Production of 48 V through the 16 O + 40 Ca Fu sion Evaporation Reaction ................................ ................................ ................................ .................. 233 Table B4 : Cross Section Data from PACE4 for the Production of 48 Cr through the 16 O + 40 Ca Fusion Evaporation Reaction ................................ ................................ ................................ ...... 234 Table B5 : Cross Section Data from PACE4 for the Production o f 52 Mn through the 16 O + 40 Ca Fusion Evaporation Reaction ................................ ................................ ................................ ...... 235 Table B6 : Cross Section Data from LisFus for the Production of 43 Sc thro ugh the 16 O + 40 Ca Fusion Evaporation Reaction ................................ ................................ ................................ ...... 236 Table B7 : Cross Section Data from LisFus for the Production of 44 Sc through the 16 O + 40 Ca Fusion Evaporation Reaction ................................ ................................ ................................ ...... 238 Table B8 : Cross Section Data from LisFus for the Production of 48 V through t he 16 O + 40 Ca Fusion Evaporation Reaction ................................ ................................ ................................ .................. 239 Tabl e B9 : Cross Section Data from LisFus for the Production of 48 Cr through the 16 O + 40 Ca Fusion Evaporation Reaction ................................ ................................ ................................ ...... 240 Table B10 : Cross Section Data from LisFus for the Production of 52 Mn throu gh the 16 O + 40 Ca Fusion Evaporation Reaction ................................ ................................ ................................ ...... 241 Table C1 : Wavelengths Used for Identification and Quantification of Stable Elements in System Water ................................ ................................ ................................ ................................ .......... 243 Table C2: ICP - OES Instrument Settings Used in Chapter 5 and 6 ................................ ............... 244 xii LIST OF FIGURES Figure 1.1: Radioconjugate Int eracting with a Receptor on the Surface of a Cell. ........................ 2 Figure 1.2: DOTA - Folate Conjugate Bound to a Radiometal ................................ .......................... 3 Figure 1.3: Simpl ified Decay Scheme for 47 Sc and its Parent 47 Ca ................................ .................. 4 Figure 1.4: Chart of the Nuclides Region around 47 Sc ................................ ................................ .... 6 Figure 1.5: De piction of Di fferent Nuclear Reactions Resulting in the Production of 47 Sc. ........... 8 Figure 1.6: Isotope Harvesting from the NSCL ................................ ................................ .............. 11 Figure 2.1: Water - B ased Target for Low Intensity Collections ................................ ..................... 17 Figure 2.2: FRIB Isotope Harvesting Target Design ................................ ................................ ....... 20 Figure 2.3: Placement of the Isoto pe Harvesting Target at the Beam Blocker Position .............. 21 Figure 2.4: Overview and C ross - Sectional View of Target ................................ ............................ 23 Figure 2.5: Ti64 Tar get Window ................................ ................................ ................................ .... 23 Figure 2.6: Schematic Drawing of Generation 2 Beam Blocker ................................ .................... 24 Figure 2.7: CT Scan of Beam Blocker Made through Additiv e Manufacturing ............................. 25 Figure 2.8: Schematic of Gen eration 1 Water System ................................ ................................ .. 27 Figure 2.9: Images of Generation 1 Target and Water System at Beam End Station .................. 27 Figure 2.10: Column Configuration for Aqueous Ion Collection ................................ ................... 29 Figure 2.11: Diagram of Isotope Harvesting Sy stem for Low Intensity 48 Ca Irradiation .............. 33 Figure 2.12 : Diagram of Isotope Harvesting System for the Proton Irradiation .......................... 34 Figure 2.13: D rawing of Ion Exchange Resin Bed ................................ ................................ ......... 36 Figure 2.14: Images of the Isotope Harvesting System for the High Intensity Proton Irradiation 38 xiii Figure 2.15: Schematic D iagram of Isotope Harvesting System for High Intensity 48 Ca Irradiation ................................ ................................ ................................ ................................ ....................... 40 Figure 2.16: Images of the Isotope Harvesting System for H igh Intensity 48 Ca Irradiat ions ........ 41 Figure 2.17: Aqueous Harvesting Loop for High Intensity 48 Ca Irradiation ................................ .. 42 Figure 3.1: Scaled Beam Current During Experimen t ................................ ................................ ... 49 Figure 3.2: Depiction of Layers Traversed by 40 Ca 20+ Beam ................................ .......................... 50 Figure 3.3: Average LET vs. Experimental G - Values ................................ ................................ ...... 55 Figure 3 .4: Measurements of the Temperature and Conductivity of the Water during the Irradiation ................................ ................................ ................................ ................................ ..... 62 Figure 3.5: Example Gamma - Ray Spectra of Sys tem Components ................................ .............. 66 Figure 3.6: Hydrogen Gas Production and Beam Current ................................ ............................ 72 Figure 3.7: Optical Image of Irradiated Ti64 Disk ................................ ................................ ......... 76 Figure 4.1: Harvesting System Overview ................................ ................................ ...................... 85 Figure 4.2: Calibr ation of Target Current Readings ................................ ................................ ...... 87 Figure 4.3: Timeline of Irradiation ................................ ................................ ................................ 88 Figure 4.4: Collection, Elution, and Preparation of Cationic Radionuclides ................................ . 91 Figure 4.5: Elution Profiles for Separation Methods 1, 2, and 3 ................................ ................ 103 Figure 4.6: 47 Ca Purification Gamma Spectra ................................ ................................ ............. 108 Figure 4.7 : St able Ions in System Water Compared to Purified 47 Ca Fractions .......................... 112 Figure 5.1: Timeline for Proton Irradiation ................................ ................................ ................. 117 Figure 5.2: Sc hematic Cross Section of the Isotope Harvesting Target ................................ ...... 118 Figure 5.3: Depiction of Recoil and Production of Radionuclides. ................................ ............. 12 4 Figu re 5.4: Depiction of Recoil Fraction Estimation ................................ ................................ ... 125 xiv Figure 5.5: Proton Beam Implanted in Isotope Harvesting Target ................................ ............. 130 Figure 5.6: A ctivity of 48 V in the System ................................ ................................ ..................... 133 Figure 5.7: Total Power and Areal Power for Various Beams ................................ ..................... 134 Figure 5.8: Linear Power Deposition o f Various Beams through the Target .............................. 139 Figure 6.1: Chemical Structure of DTPA - TOC ................................ ................................ .............. 143 Figure 6.2: Beam Structure and Sample Collecti on ................................ ................................ .... 147 Figure 6.3: Separation Scheme for the Purification of 47 Ca ................................ ........................ 153 Figure 6.4: 47 Ca/ 47 Sc Generator Activity ................................ ................................ ..................... 154 Figure 6.5: Modified Pseudo Generator Procedure ................................ ................................ ... 155 Figure 6.6 : TLC Quality Control Test ................................ ................................ ........................... 160 Figure 6.7: Collection Resin Bed Elution Profiles ................................ ................................ ........ 165 Figure 6.8: Elution Profile for Purification of 47 Ca ................................ ................................ ...... 167 Figure 6.9: Elu tion Profile of 47 Sc from DGA 2 ................................ ................................ ............ 168 Figure 6.1 0: Elution from DGA 1 in Generator 4 ................................ ................................ ......... 170 Figure 6.11: Phosphor Images for Thin Layer C hromatography Quality Control Tests .............. 173 Figure 7.1: Experimental Beam Structure ................................ ................................ ................... 181 Figure 7.2 : Schematic Description of the 47 Ca/ 47 Sc Generator Procedure ................................ . 185 Figure 7.3: Stable Ca 2+ /K + Separation ................................ ................................ ......................... 188 Figure 7.4: Separati on Elution Profiles for AG MP - 50/HCl Separa tions ................................ ..... 189 Figure 7.5: Elution Profile for DGA 2 in Generators 1 and 3 ................................ ....................... 191 Figure 7.6: Phosphor Images of Thin Layer Chromatography Q uality Control Tests ................. 193 Figure 8.1: Schematic of Main Components of Water System ................................ ................... 199 xv Figure 8.2: Half - Lives for 47 Ca: LSC Count Rat es ................................ ................................ ......... 209 Figure 8 .3: Half - Lives for 47 Ca and 47 Sc: Gamma - Ray Spectroscopic Peaks ................................ 211 Figure 8.4: Branching Ratios for Three Most Inten se 47 Ca Gamma Ray s ................................ ... 212 Figure 8.5: Decay Scheme of 47 Ca Beta Decay to 47 Sc ................................ ................................ 215 Figure D1: Linear Trends Used to Interpolate Geometry Correction Factors 248 1 Chapter 1 : Introduction As the second leading cause of death in the United States, cancer claimed 595,930 lives in 2015. [1] About 90% of these deaths ar e caused by metastatic cancer, or cancer that has spread to other parts of the body. [2] These diseases have many treatment options including surgery, chemotherapy, and radiotherapy. Large tumors ( i.e. , >1 g) can be treated using surgery and external radiotherapy since th ey can be readily identified. Some m icrometastatic t umors are too small to be imaged accurately, limiting treatment options for these small growths. Targeted internal radiotherapy is one possible treatment option for small metastatic a nd micrometastatic tu mors, usi ng the targeted delivery of radionuclides t o treat subclinical growths throughout the body. Numerous radionuclides including alpha, beta, and Auger electron emitters are being considered for use in targeted internal radiother apy. Radionuclides em itting th ese forms of radiation ca n damage the DNA o f a cell and prompt apoptosis, or cell death. This occurs when the radiation directly deposits energy in the DNA or produces chemical species, such as radicals that can secondarily in teract with the DNA. When thes e measures are successful i n inducing a dou ble strand DNA break, they often lead to irreparable damage and apoptosis. Additionally, radionuclides that emit positrons and low energy gamma rays ( i.e. , 100 keV - 400 keV) can be use d to image metastatic cancerou s growths in a targeted m an ner with Positro n Emission Tomography (PET) and Single Photon Emission Computed Tomography (SPECT) imaging, 2 Figure 1 . 1 : Radioconjugate Interacting w ith a Receptor on the Surface of a Cell. respectively. T ogether, these c apabilities would allow for metastatic cancer to be diagnosed and treated using targeted radionuclides. To target cancerous growths, the radionuclides are bound to a targeting structu re composed of a biol ogica lly active and a chelating comp onent. The biolo gical portion of the vector directs the movement of the radiopharmaceutical in the body, binding to the surface of cells upon interaction with receptors. Some radiopharmaceuticals are absorbed by targeted cells, d ecreasing the distance and increasing the p otential for interactions between the emitted radiation and the cell DNA. Folate is one example of a small biological molecule that binds to folate receptors which are often over - expr essed in a variety of cancers, including ovarian and lung cancer. [3,4] Attached t o the biological entity is a chelating structure that binds to a radiometal. DOTA ( 1,4,7,10 - t etraazacyclododecane - 1,4,7,10 - tetraacetic acid ) is a macrocy clic chelating ligand that is known to bind to a variety of metals with its four tertiary amines and 3 Figure 1 . 2 : DOTA - F olate C onjugate B ound to a Radiometal [5] The portion of the molecule highlighted on the top left is the b iological folate component that binds to folate receptors; the lower portion of the molecule that is highlighted is the DOTA chelating ligand bound to a radiometal such as scandium, shown here as M . 1 four carboxylic acid groups. This process is shown sc hematically in Figure 1 and an example of a DOTA - folate conjugate is given in Figure 2 (modified from [5] ). mean ing the radionuclides are isotopes of the same element. These radionuclides then have identical chemistry fo r radiolabeling and interacting within the body. [6] An initial low - dose imaging step can demonstrate the biodistribution a nd dosimetry of this chemical entity in the body. This can confidently be followed by high - dose therapy, since the therapeutic radiocon jugate will behave identically to the imaging agent in the body. [7] One element that has such a theranostic pair is scandium. The - emitting 47 Sc has a 3.35 day half - life and a relatively low ene rgy emission at an average energy of 162 keV (E = 142.6 keV, I = 68.4%; 1 This figure was originally published in JNM . Cristina Müller, Maruta Bunka, Stephanie Haller, Ulli Köster, Viola Groe hn, Peter Bernhardt, Nicholas van der Meulen, Andreas Türler, and Rog er Schibli. Pr omising Prospects for 44 Sc - / 47 Sc - Based Theragnostics: Application of 47 Sc for Radionuclide Tumor Therapy in Mice. J Nucl Med. 2014; 55; 1658 - 1664. © SNMMI. 4 E = 203.9 keV, I = 31.6%) . [8] This low energy - particle is suited to tr eat small cancer metastases with a tumor diameter in the range of a few millimeters ( i.e. , 2.0 - 3.8 mm) . [9] This decay is also ac companied by a 159 keV gamm a ray (I = 68.4%) which is low enough energy to be v isualized with SPECT imaging . [8] The se decay properties as well as properties of 47 Ca, the parent of 47 Sc, are given in Figure 3. As a match to this therapeutic scandium isotope, both 43 Sc and 44 Sc have + emissions that are suitable for PET imaging. [10 13] These isotopes have similar decay properties with low energy + emission s and half - lives of approximately 4 hours. [14,15] However, 44 Sc has a high energy gamma ray at 115 7 keV (I = 99.9%), making 43 Sc mor e suitable for radiotherapy . [15] These theranostic isotopes would be an improvement over the 68 Ga/ 177 Lu pair that is routinely used in the clinic for PET imaging and therapy, respectively. Since these isotopes are not of the s ame element, they bind to the targeting vector differently. [6] Differential binding Figure 1 . 3 : S implified Decay S cheme for 47 Sc and its Parent 47 Ca 2 [8] 2 Reprinted from ACS Omega, 5; E. Paige Abel, Katharin a Domnanich, Hannah K. Clause, C olton Kalman, Wes Walker, Jennifer A. Shusterman, John Greene, Matthew Gott, and Gregory W. Severin; Production, Collection, and Purification of 47 Ca for the Generation of 47 Sc through Is otope Harvesting at the National Supe rconducting Cyclotron Laboratory , 27864 - 27872, Copyright (2020), with permission from the American Chemical Society under an ACS AuthorChoice License. 5 may affect as pects such as the pharmacokinetics of the two radiopharmaceuticals and binding with the targeted receptors. This means that the therapeutic radionuclide may have a different biodistribution than the diagnostic rad ionuclide, leading to lower doses for the t umorous tissue and higher doses for normal or radiosensitive tissues than was expected. In addition, the half - lives of the two radionuclides differ by more than two orders of magnitude with 68 Ga having a half - life of 68 minutes compared to 177 Lu with a hal f - life of 6.65 days. [16,17] After a few hours in the body, 68 Ga no longer provides information about the biodistribution and dosimetry of 177 Lu. [18] Thus, the chemical and nuclear properties of the scandium isotopes are much better matche d to accomplish paired therapy and diagnosis. In 2014, a study was performed by M ü ller et. al. to demonstrate the therapeutic ability of 47 Sc. [5] Several studies were performed including in vivo tests with a 47 Sc - D OTA - folate radioconjugate to take SPECT/CT images as well as to test the therapeutic effec tiveness of 47 Sc. The radioconjugate was well localized in the cancerous tissues, allow ing for excellent visualization of the tumors. The high resolution afforded by this radioconjugate would be helpful for pretherapy dosimetry. The therapeutic studies per formed in vivo were marked by delayed tumor growth and therefore, a 54% increased survival time among the treated vs. control mice. This potential radiopharmaceutical demonstrate d antitumor properties comparable to that of the 177 Lu - DOTA - folate radioconju gate. [5] Even though these scandium isotopes have attracted research attention, a sustainable production route for 47 Sc has not yet been established . Several methods under study have encountered many of the same diff iculties: prohibitively expensive target material, the formation of other scandium isotopes, or low production yields of 47 Sc. [10,19 26] These target 6 materials and scandium isotopes are shown in the relevant portion of the C hart of the N uclides in Figure 4. The high c ost of these production routes comes from the necessity of using isotopically enriched Ca and Ti target materials to increase the relative abundance of 46 Ca, 48 Ca, or 47 Ti in the target . The calcium targets in particular are very expensive due to their lo w natural abundances of 46 Ca and 48 Ca ( i.e. , 0.004% for 46 Ca and 0.187% for 48 Ca ) . Additionally, 44m, 46, 48 Sc are often produce d as impurities even when an optimized beam energy range is used to produc e 47 Sc. The presence of these scandium impurities could add significantly to the dose received by a patient without any added therapeutic benefit due to their gamma - ray emissions, as 46,48 Sc each have two high energy and intensity gamma rays ( i.e. , 46 Sc: 889.3 keV at 99.98 and 1120.5 keV at 99.99% and 48 Sc: 98 3.5 keV and 1312 keV both at 100%) and 44m Sc decays to its Figure 1 . 4 : Chart of the Nuclides Region around 47 Sc For the squares representing radionuclides, the half - life is given , while for the squares re presenting stable species, the natural abundance is given. 3 3 Reprinted from ACS Omega, 5; E. Paige Abel, Kathar ina Domnanich, H annah K. Clause, Colton Kalman, Wes W alker, Jennifer A. Shusterman, J ohn Greene, Matthew Gott, and Gregory W. Severin; Production, Collection, and Purification of 47 Ca for the Generation of 47 Sc through Isotope Harvesting at the National Su perconducting Cy clotron Laboratory, 27864 - 27872, Copy right (2020), with permission fr om the American Chemical Society under an ACS AuthorChoice License. 7 daughter 44 Sc which has an emission of 1157 keV at an intensity of 99.9%. [15,27,28] S ince all three of these isotopes are of the same element, they cannot be separated chemically to produce 47 Sc with a higher radionuclidic purity. Each of these radionuclides is also sufficiently long lived ( i.e . , t 1/2 = 2.44 d, 83.79 d, and 43.67 h for 44m,46,48 Sc, respectively), making it is impractical to wait for them to decay to significantly improve the radionuclidic purity of the 47 Sc product. [15,27,28] Therefore, any method that produces a sizeable 44m Sc, 46 Sc, or 48 Sc impurity is n ot a viable production method. Proton and neutron irradiations as well as photonuclear reactions have been studied for the production of 47 Sc. [10,19 26] Several of these reactions are depicted in Figure 5 as well as described below. One production route investigated uses cyclotron irradiation of a natural calcium target or a 48 Ca en r iched target to induce the following reaction: 48 Ca(p,2n) 47 Sc. [19] Both 46 Sc and 48 Sc are produced in such irradiation s by either the 48 Ca(p,3n) 46 Sc or the 48 Ca(p,n) 48 Sc reaction. In the energy range that most medical cyclotrons operate (< 20 MeV), the activity of 48 Sc produced is significant compared to that of 47 Sc ( i.e . , 13% of the scandium activity depending on the energy range used) . The quantity of 47 Sc produced from a natural calcium target is also quite low due to the low natural abundance of the target nucleus, 48 Ca. By increasing this abundance with an enriched 48 Ca target, the production yield for 47 Sc would increase. However, the production yield of 48 Sc as well as the production cost would also increase with such a target. Another possible production route through proton irradiation uses a titanium target. Natural Ti targets have been employed in several studies, utilizing the following reactions to 8 Figure 1 . 5 : Depiction of D ifferent N uclear R eactions R esulting in the P roduction of 47 Sc. Top row: 48 Ti(p,2p) 47 Sc, 49 Ti(p, 3 He) 47 Sc, 50 47 Sc, 47 Ti(n,p) 47 Sc, and 46 Ca(n, 47 Sc. Bottom row: 48 Ca(p,2n) 47 Sc and 48 47 47 Sc. produce 47 S c: 48 Ti(p,2p), 49 Ti(p, 3 He), and 50 Ti(p, ). [20,21] The optimal energy ran ge for producing 47 Sc and minimizing 46,48 Sc production is 22 - 33 MeV, which again is higher beam energy than most medical cyclotrons can provide. [20] The 46,48 Sc impurity is re latively low at approximately 3% of the activity produced. However, 44m Sc is a major contaminant, producing comparable activities to 47 Sc at all energies used with this production route . It h as b een suggested that using 50 Ti targets at proton energies of <20 MeV may allow for acceptable production yields for 47 Sc despite the low cross section observed in this energy range with a natural Ti target. [20] This enriched target may als o reduce the coproduction of 44m Sc which is primarily produced through the 47 Ti (p, ) and 48 Ti (p,n ) reactions. Work has also been done with an enriched 48 Ti target to reduce the 44m ,48 Sc con tami nant s , but this method increase s the 46 Sc impuri ties produced and has only been explored at energies > 48 MeV . [21] 9 While the higher energy proton spallation reactions create many products, lower energy neutron reactions can be more finely controlled to favor only one nuclide. Neutron irradiations that produce 47 Sc involve the us e of fast neutrons in the 47 Ti(n,p) 47 Sc r eaction or thermal neutrons in the 46 Ca(n, ) 47 47 Sc reaction. [10,21] Both of these reactions require enriched target materials to reduce contaminants . The fast neutron reaction of ti tani um requires very high neutron fluxes o n the order of 10 14 n/(cm 2 s) to produce activities in the GBq range from a small target. [10] Few reactors can produce a neutron flux t h is large, limiting the practicality of this production route. The 46 Sc impurity from this reaction also limits its use with up to a 11.5% impurity being produced in some neutron spectra. [10] The thermal neutron reaction on an enriched 46 Ca target can produce GBq activities of 47 Sc with a several day irradiation at a thermal neutron flux of 10 14 - 10 15 n/(cm 2 s). This production route avoid s significant scandium impurities by using both an enriched target and the production of a 47 Ca/ 47 Sc generator. [10] The primary disadvanta ge of this production route is the requirement of expensive Ca target material enriched in 46 Ca to achieve a high yield and radionuclidic purity for 47 Sc . Two photonuclear reactions have also been explored: 48 Ca( ,n) 47 Ca 47 Sc and 48 Ti( ,p) 47 Sc. [22 26] In these reactions, high energy electrons are generated with an electron linear accelerator and are some of the kinetic energy of the electrons into bremsstrahlung. The photons then pass through a target material to produce the desired radionuclide. The first photonuclear reaction listed above uses an enriched 48 Ca target due to the low abundance of this isotope in a natural Ca target. Even with t his enriched target, however, activities of 47 Sc only reach the low MBq g - 1 range using a 1 mA beam current . [22] No scandium impurities were detected using this production route as there 10 are no significant photonuclear reactions with a calcium target that lead to the production or generation of 44m,46,48 Sc. The second p h otonuclear reaction mentioned above uses an enriched 48 Ti target. [23,24] Impurities from 46 Sc can be controlled by using an enriched target and by keeping the maximum photon energy below the energy threshold of the 48 Ti( ,pn) 46 Sc reaction ( i.e. , < 22 MeV) . With optimized parameters, this production route produced a 47 Sc yield of hundreds of MBq g - 1 . It has been predicted but not yet experimentally verified that both of these photonuclear reactions can produce higher 47 Sc yield s wit h optimized tar get geomet ries for the distribution of bremsstrahlung rays and higher beam currents . [22,24] The 48 Ti( ,p) 47 Sc production method was first studied in 1977 , but has recently gained more attention as a potential sustainable production method for a 47 Sc supply. Recent studies by Rotsch et . al . in 201 8 and Loveless et . al . in 2019 have demonstrated successful purification of 47 Sc produced in this way. [25,26] Additionally, the high purity 47 Sc sample produced in the Loveless et al study was us e d to radiolabel DOTA - TOC, demonstrating the feasibility of using 47 Sc from this production method in preclinical studies . [25] While research into a future sustainable production method for 47 Sc is ongoing, an untapped supp l y of 47 Sc exists at nuclear physics accelerator facilities that provide fast, heavy ion beams, such as the National Superconducting Cyclotron Laboratory (NSCL) and in the future at the Facility for Rare Isotope Beams (FRIB). The NSCL produces high energy p rimary , stable beam to produce a user - specified secondary , radioactive beams through projectile fragmentation reactions . While a single isotopic secondary beam is generally used by the experimenters, the unreacted primary beam is collected in a solid meta l beam blocker. 11 Although the amount of primary beam that passes unreacted through the target depends on the secondary beam specified by the users, often over 90% of the primary beam i s unreacted. Using LISE++, a program that allows for optimization of expe r imental parameters for fragmentation reactions in a thin target and the separation of the resulting radioactive fragments in a zero - degree spectrometer, the amount of unreacted primary beam transmitted through the target can be estimated for a particular s econdary beam. [29] For example, a n experiment requiring the production of a 40 Mg secondary beam is predicted to have 94.8% of the primary 48 Ca beam left unreacted after the target . Figure 1 . 6 : Isotope Harvesting from the NSCL A stable, primary accelerated beam is used to produce exotic, radioactive secondary be a ms at the NSCL. A large portion of the initial accelerated stable beam goes unreacted at the target, is separated out of the radioactive cocktail beam at the first dipole magnet in the A1900 fragment separator and i s then stopped in a beam stopper. A flow i ng - water target system can replace the solid metal beam blocker at the NSCL to allow for isotope harvesting. 4 4 Reprinted from Nuclear Instruments and Methods in Physics Research B, 478; E. Paige Abel, Katharina D omnanich, Colton Kalman, Wes Walker, Jonathan W. Engl e, Todd E. Barnhart, Greg Severi n; Durability test of a flowing - water target for isotope harvesting, 34 - 45, Copyright (2020), with permission from Elsevier. 12 This surplus primary beam can be collected off axis in a thick water target to induce further nuclear r eactions in tandem with the primary exper i ment to produce longer - lived radionuclides. This emerging production method at accelerator facilities has been coined [30] The harvesting water target can replace the solid metal beam blocker presently in the beam line after the first dipole magnet in the A1900 fragment separator at the NSCL (see Chapter 1 Figure 1.7). [31] At this point in the facility, the remaining primary beam after the fragmentation target has been separated from t he radioactive beam cocktail and can be use d for isotope harvesting. Figure 6 depicts the path of the primary beam as it produces exotic fragments at the fragmentation target, is separated from these exotic fragments in the first dipole magnet of the A1900 fragment separator, and is sent into a flo wing - water target for isotope harvesting. The 48 Ca primary beam is the second most requested primary beam at the NSCL . When the unreacted portion of this beam is implanted in a water target, 47 Ca is the most abun dantly produced fragment. In the example of the 40 Mg secondary beam, the unused 48 Ca beam would produce 7.98(8) x 10 9 pps of 47 Ca (see Chapter 5 Section 5.2.3 and 5.3.1 for the measurement of this production rate). For a 120 - hour experiment (the typical le ngth of a nuclear physics experiment at the NSCL), this would result in 115(1) mCi or 4.27(4) GBq of 47 Ca produced in the isotope harvesting water target. Even with a 24 hour wait ing period to allow for short - lived activity to decay, almost 100 mCi of 47 Ca would remain. In the future with FRIB, the production rate per particle should be great er than that with the 48 Ca beam from the NSCL due to the higher energy beam ( i.e. , 240 MeV/nuc leon at FRIB vs. 140 MeV/nucleon at the NSCL). With the anticipated FRIB 4 8 Ca primary beam intensity and the 47 Ca production rate measured at the 13 NSCL, a t least 3.7 x 10 12 pps will be produced at FRIB. This rate would result in the production of > 14 Ci or 520 GBq during a 24 - hour experiment. In addition to 47 Ca, other calcium and scandium isotopes including 45 Ca and 44m,46, 47, 48 Sc are produced in the water target. As with other production methods, these co - produced scandium isotopes cannot be chemically removed from the 47 Sc product and could add significant dose burden if us ed in therapeutic applications. Instead of harvesting 47 Sc directly, 47 Ca can be harvested and used to generate a high purity sample of 47 Sc. This avoids the many scandium isotopes t hat are co - produced in the water. While 45 Ca is co produced and will not be separated from the 47 Ca, it decays to stable 45 Sc. This means that the presence of 45 Ca in the 47 Ca/ 47 Sc generator should not interfere with the radionuclidic purity of the generate d 47 Sc. Additionally, the small number of atoms of 45 Ca produced and the l ong half - life of this radionuclide indicate s that it will not affect the specific activity of the 47 Sc sample. This approach also allows for the processing of Ca 2+ ions which ha ve simple chemistry across a wide range of acidity and oxidative potentials as opposed to scandium ions which are easily hydrolyzed in neutral and basic conditions. Therefore, using a 47 Ca/ 47 Sc generator w ould allow for production of high - quality samples o f 47 Sc. Several experiments have been performed in the past and in the work pre sented here to demonstrate the feasibility of isotope harvesting and optimize techniques for collection and purification of isotope harvested radionuclides for a variety of appl ications. [32 37] The first proof of concept study was performed at the NSCL with an purified, fast 24 Na secondary beam collected in a small water volume. This experiment successfully demonstrated the ability to collect kBq qua ntities of an 85 MeV/nucleus 24 Na beam. [32] Other experiments demonstrated 14 successful collection and purification of a 67 Cu secondary beam from a mixture of other co - produced radionuclides in an aqueous medium . [34 ,38] In one experiment, the secondary beam contained only 2.9% 67 Cu among se veral other radionuclides. This experiment successfully recovered and purified a significant portion of the collected 67 Cu with a high radionuclidic purity. The purified product was then used to radiolabel a clinically significant antibody , whic h was used to demonstrate tumor uptake in a n in vivo study. These experiments clearly demonstrated on a small scale the feasibility of using isotope harvesting as a production method for radionuclides with applications in nuclear medicine. Moving forward w ith this is otope production method requires experiments that more closely resemble the anticipated isotope harvesting conditions at FRIB. This include s increasing the water volume, using a flowing - water target and isotope harvesting water system, and devel oping techn iques for online collection of radionuclides from the system water. Additionally, these experiments should be carried out with higher intensity primary beam s to produce radionuclides in the target compared to the low intensity radioactive second ary beams t hat were collected in a water target previously. The experiments presented here have allowed for the development of techniques and equipment that culminated in an 8 hour irradiation of a 50 L flowing - water isotope harvesting system with a 140 Me V/nucleon 8 0 pnA 48 Ca beam that produced a mCi level 47 Ca/ 47 Sc generator. A portion of the purified 47 Ca was also used to measure the three most intensity branching ratios for this radionuclide, reducing the uncertainty compared to values in the literature and improv ing the precision with which this product was quantified . These experiments allowed for the optimiz ation of chemical techniques to produce high purity 47 Sc that could be used in preclinical studies. The techniques and 15 knowledge gained through th ese experim ents will assist in the translation of these methods to isotope harvesting at FRIB to produce a supply of 47 Sc for further research. 16 Chapter 2 : Development of Isotope Harvesting System To capitalize on the availability of unused accelerated pr imary beam to produce long - lived radionuclides of interest for applications, an isotope harvesting target and water system were developed. These efforts began when collaborators built and tested a s tationary 100 mL water target that was irradiated with a s econdary radioactive beam. The present efforts by the Isotope Harvesting group at Michigan State University started with bridging the gap between this small, static volume used in early exploratory work and the design of the much larger flowing - water beam dump at the Facility for Rare Isotope Beams (FRIB) which is presently under construction. Several generations of isotope harvesting targets and water systems have been developed and tested through i rradiations of several stable, accelerated primary beams a t the National Superconducting Cyclotron Laboratory. The targets and water systems developed will be presented here. 2 .1 Introduction Previous isotope harvesting experiments at the NSCL have used a water - based target system and a secondary radioactive b eam for proof - of - concept experiments for collection in and extraction from a water medium [32 35] . This target system included a 100 mL static water target enclosed in a PTFE (polytetrafluoroethylene) body with a thin Kapton front window shown schematically in Figure 2 .1 [32] . Harvesting experiments with this target collected low intensity particle picoamperes (ppA) radioactive beams produced at the NSCL. Experiments with radioactive beams purified through separation in the A1900 fragment separator [31,33] , 17 cocktail beams containing a mix of secondary radioactive fragments [32,34] , and cocktail beam s containing both a variety of radioactive particles and stable primary beam particles [35] were carried out. The beam particles passed through the Kapton front window and were stopped in the interior water volume demonstrating successful collection of fast, heavy ions in an aqueous medium. Sepa ration method s performed on the collection samples also supported the feasibility of purifying the collected radioactive products of interest for use in further applications [33 35] . Figure 2 . 1 : Water - Based Target for Low Intensity Collections A vertical cross section of the water - filled target encased in PTFE and a Kapton front window is shown schematically. This target was used to colle ct fast radioa ctive secondary beams. The target was filled with water remotely through the top left port and was drained remotely with the port at the back of the target (on the left in the drawing). 5 5 Reprinted from Nuclear Instruments and Metho ds in Physics Re search A, 747; Aranh Pen, Tara Mastre n, Graham F. Peaslee, Kelly Petr asky, Paul A. DeYoung, David J. Morrissey, Suzanne E. Lapi ; Design and construction of a water target system for harvesting radioisotopes at the National Superconducting C yclotron Laboratory , 62 - 68, Copyright (2014), with permission from Elsevier. 18 Given that the full beam intensities at the NSCL ar e already on th e order of particle nanoamperes (pnA) and the anticipated intensities at FRIB will be on the order of particle microamperes (p A), the target system for isotope harvesting of the primary beam at these facilities will need to withstand with m any orders of magnitude higher beam intensities than those tested with the static water - filled target system. Additionally, these heavy ion b eams will be accelerated to higher energies at FRIB compared to those available at the NSCL and therefore, will del iver more power to the water. Together, these effects will result in primary beam powers of up to 400 kW with up to 325 kW deposited in the i sotope harvesting beam dump at FRIB. These power levels are enormous compared to the maximum power depositions of 9 x 10 - 6 k W used in the previous static water target irradiations and 0.5 kW with the available beams at the NSCL. A direct result of high be am power in the target is a large heat deposition as the beam stops and transfers kinetic energy to the target shel l and water interior. One particular concern is the heat transfer from the target shell to the water interior. Insufficient heat transfer and dissipation can lead to temperatures in the target shell significantly above the boiling point of the system water , resulting in nucleate boiling at the interface between the target shell and the water. This produces vapor pockets at the interface of thes e materials further reducing heat transfer. If an irradiation were to continue under these circumstances, the targe t shell could reach extreme temperatures and potentially beyond the melting point of the material. Another effect of high - power beams is the production of corrosive radiolysis products, such as radical and oxidizing species, in the water interior of the t arget. Exposure of target materials to this environment could lead to degradation and mechanical weakening of the target. [39] 19 Both the high temperature and corrosive environment produced with a high - power deposition could result in catastrophic failure of the target. This presents a serious safety issue if a leak in th e system should develop, releasing high levels of radioactivity transported in the flow of water. Additionally, failure of the beam dump would interrupt the nuclear physics program which is the primary mission of accelerator facilities like the NSCL and FR I B. To avoid target failure and interruptions to the experimental program at FRIB, the beam dump must be able to absorb the high - power primary beam in a safe manner for at least 1 year or 5500 hours of operation. Thus, the design for the isotope harvesting target at FRIB must be able to minimize and withstand damaging effects from this large power deposition a target. The current design plan for the isotope harvesting target or beam dump at FRIB has been described in Avilov, et al [39] . The design features a rotating cylinder with a Ti64 alloy ( grade 5 alloy: 6% Al, 4%V , mass balanced with Ti) shell and a flowing - water interior depicted in Figure 2 .2. The target shell material of Ti64 was chosen for many reasons such as its strength and resistance to corrosion as well as its availability as a material that can be used in additive manufacturing techniques such as 3D printing . The target is a torus 70 cm in diameter and 8 cm tall along t h e rim that will be irradiated. To effectively dissipate the heat produced from the primary beam power deposition, the target will be rotate d at 600 rpm and water will flow through the interior at a rate of 60 gallons per minutes. The target shell also inc l udes a double wall depicted in Figure 2 .2b, that directs the flow of water into the target through a narrow 2 mm space between two thin 0.5 mm Ti64 walls. This dramatically increases the flow of water over the target shell and therefore, the heat transfer between the target shell and the water. 20 Figure 2 . 2 : FRIB Isotope Harvesting Target Design A cross sectional view of the isotope harvesting beam dump designed for FRIB (a) and a closer view of the double w alled design of the target shell along the bottom and rim of the beam dump target (b). In both pictures, the blue arrow depicts the flow of cold water entering the beam dump and the red arrow traces the path of the higher temperature water leaving the bea m dump. The picture on the left also has a green arrow that indicates the direction of the rotation of the target and an orange arrow that indicates the direction of the primary beam. 6 All of these features facilitate the removal of heat from the target sh ell and prevent boiling at the interface, as described previously. In addition to using a corrosion resistant material for the target shell, a water system is needed to monitor and red uce the effect of the corrosive products of radiolysis in the target. T he p resent efforts focused on the long - lived molecular products of radiolysis in prototype isotope harvesting systems. Multiple sensors have been used during experiments to understand t he buildup of these products as well as anticipate unsafe levels before the y occur. Components were also tested to destroy or purge the system of these radiolysis products to prevent a concentrated build up. Finally, the prototype water system must contain components that 6 Reprinted from Nuclear Instruments and Methods in Physics Research B, 376; Mikhail Avilov, Adam Aaron, Aida Amroussia, Wladimir Bergez, Carl Boehlert, Thomas Burgess, Adam Carro ll, Catherine Colin , Florent Durantel, Paride Ferrante, Tiffany Fourmeau , Van Graves, Clara Grygiel, Jacob Kramer, Wolfgang Mittig, Isabelle Monnet, Harsh Patel, Frederique Pellemoine, Reginald Ronningen, Mike Schein ; Thermal, mechanical and fluid flow asp ects of the high po wer beam dump for FRIB , 24 - 27, Copyright (2016), with permi ssion from Elsevier. 21 carryout the main purpose of isotope harvesting: the e xtra ction of radionuclides of interest for further use in applications. Isotope harvesting efforts at the NSCL have focused on the design and testing of targets and water systems for i sotope harvesting at increasingly FRIB - like conditions. In the followin g ch apters, several isotope harvesting experiments that have been carried out to test the durability of these systems (Chapters 3 and 5 ), and to demonstrate proof - of - concept for the ext raction of radionuclides for further applications (Chapters 4 , 6 , and 7 ) wi ll be described and the results analyzed . [37,40,41] The flowing - water target was used in three locations for the present experiments : (1) at the end station of a beam line at the NSCL for a low intensity experiment with 48 Ca (Chapter 4 ), (2) at the end of the beam line at the Cyclotron Laboratory at the University of Wisconsin - Madison for a high intensity proton irradiation (Chapter 5 ), and (3) at the beam blocker position in the A1900 fragment separator at the NSCL for two high intensity 48 Ca irradiations (Chapter 6 , 7 , and 8 ). The beam blocker position is indicated in a Figure 2 . 3 : Placement of the Isotope Harvesting Target at the Beam Blocker Position The solid metal beam blocker is po sitioned after the first dipole magnet in the A1900 Fragment Separator at the NSCL and was temporarily replaced with the isotope harvesting target for a few high - intensity isotope - harvesting experiments [31] . 22 schematic of the cyclotrons and the A1900 Fragment S eparator in Figure 2 . 3 . The isotope harvesting targets and water systems used in the present experi ments are introduced below. 2 .2 Target Design 2 .2.1 Generation 1: Low Intensity 40 Ca Irradiation The first flowing water target for isotope harvesting was us ed to test the durability of a disk of Ti64 material and to demonstrate the production and collecti on of radionuclides produced from a primary beam irradiation of a water target (see Chapter 3 ). The target used for this experiment had a shell made mostly o f high - density polyethylene . The shell of the target was designed for a flow of water to enter at a n inlet in the bottom and exit through an outlet at the top. The internal chamber could hold approximately 120 mL of water at a time. The target and the plac ement at the end of the beam line are shown in Figure 2 . 4 . At the front of the target, a small Ti64 disk was used as the beam entrance window. This disk, shown in Figure 2 . 5 , was 0.7 mm thick with a diameter of 15 mm exposed to the water. The surface of th e disk was divided into two halves, with one half milled smooth and the other half left with the ro ugh texture from the manufacturing process. The material was fabricated by Stratasys through 3D printing since this manufacturing technique which will be used to make the FRIB beam dump. [39] This technique can fabricate the complex internal design in a single process that traditional machining cannot achieve. Even though only a simp le disk of the material was used in this target, this technique was used so that the material was made in an analogous way to the material to be used for the FRIB beam dump. Therefore, this target design moved closer to the anticipated FRIB beam dump desig n with the Ti64 front face and the flowing - water interior. 23 During the irradiation, a leak developed in the first - generation target between the main shell of the target and the back plate (Figure 2 . 4 b). The target was not designed to be able to handle the periodic pressurization from the pulsing of the diaphragm pump that was used to flow water through th e target. This put pressure on the backplate, deformed the O - ring that was intended to make a water - tight seal, and led to a small water leak. This target design was not used for any future irradiations due to the potential radiation safety hazard it prese nted. Figure 2 . 4 : Overview and Cross - Sectional View of Target Drawing of the placement of the flowing - water target at the end of a beam line at the NSCL (a) and a cross - sectional view of the inside of the t arget where beam enters through the front Ti64 disk as shown by the teal arrow (b). Water circulated through the target entering from the bottom and leavi ng through the top as shown by the blue and red arrows, respectively. Figure 2 . 5 : Ti64 Target Window The Ti64 disk is shown inserted in the high - density polyethylene target body. The area shown was exposed to the irradiated water on the inside of the target during the experiment. The half towards the to p has the rough texture resulting from the additive m anufacturing process while the bottom half has a milled smooth texture. 24 2 .2.2 Generation 2: 1 H and 48 Ca Irradiations A next - generation flowing - water target was designed to resemble the FRIB beam dump m ore closely. This target design was used in the prot on irradiation described in Chapter 5 as well as the three 48 Ca irradiations described in Chapters 4 , 6 , 7 , and 8 . The second generation target has a shell made of Ti64 with a hollow interior and an addi t ional internal Ti64 wall close to the front face of the target. As with the FRIB target design, this double wall featured two 0.5 mm Ti64 walls with a 2 mm gap between them through which the cold incoming water flows. The water flow is then directed into a larger chamber behind the interior wall and out an exit channel. The target was operated with a water flow of approximately 11 L/min. The target body is 12.5 x 5.1 x 5.1 cm as shown in Figure 2 . 6 wit h an interior space that can hold approximately 86 mL o f water. The dimensions of this target were chosen to match the size of the existing Figure 2 . 6 : Schematic Drawing of Generation 2 Beam Blocker A vertical (a) and horizontal (b) cross section of the inside of the target is depicted. The teal arrows show that the primary beam passes through while the interior wall is the second layer of Ti64 traverse d by the primary beam. 25 beam blocker of the A1900 separator at the NSCL. This allowed the flowing - water target to replace the existing solid metal beam blocker temporarily with relatively few adjustments. The final target design was produced by 3D printing a nd therefore, had a rough texture as described p reviously. Since the inside of the target and the uniformity of the material cannot be directly observed, CT scans were taken of several targets using a micro CT scanner (Perkin Elmer, Quantum GX CT). Example CT images from scans of two targets are shown in Figure 2 . 7 . The target shown in Figure 2 . 7 a demonstrates the inconsistencies that can occur during the manufacturing process with an air bubble in the front wall and a bend in the interior wall . Air pockets in the target shell will interfere with heat transfer between the front target face and the water. Both air pockets and bends in the walls can also compromise the strength of the thin structures. In contrast, the target shown in Figure 2 . 7 b s hows uniformity in both the exterior Figure 2 . 7 : CT Scan of Beam Blocker Made through Additive Manufacturing In the first image, an air gap in the front wall (1) and a bend in the inner wall (2) can be obs erved (a). The second image shows two straight and uniform walls (b). 26 and interior wall that is desired in the construction of these targets. Only non - defective targets such as the one shown in Figure 2 . 7 b were used for irradiations. These scans also pro vide d a precise measurement of the thickness of th e walls. This measurement was used to calculate the energy of the beam in the water layer of the target which can aid in understanding the production of radionuclides and radiolysis products in the target. 2 . 3 Water System Design 2 .3.1 Generation 1: Low Intensity 40 Ca Irradiation 2 .3.1.1 Water System The flowing water system used in the low intensity 40 Ca irradiation (see Chapter 3 ) included sensors, membrane contactors, pump s , a water reservoir, ion excha nge resins, and pieces of a platinized titanium mesh arranged according to the schematic shown in Figure 2 .8 and in the pictures in Figure 2 high - density polyethylene tubes and polypropylene f ittings through which wate r was pumped at a flow rate of 1 L/min using an air operated diaphragm pump (Yamada NDP - 5FVT). Three sensors were used to detect the level of radiolysis products in the water: a Honeywell Midas Gas Detector for detecting hydrogen gas (MIDAS - E - H2X), a Mettler - Toledo Thornton pHure Sensor (2003i - UPW/120), and a Mettler - Toledo Thornton High Performance Dissolved Oxygen Sensor (InPro6050 polarographic oxygen sensor). A fourth sensor, a Mettler - Toledo Thornton UniCond Conductivity Senso - 1 sensor, part number 58 031 406) , was used to detect dissolved ions in the water which could indicate the introduction of ions to the water from corrosion. An Omega Micro - Fl o paddle wheel flow meter (part number FTB323D) was u sed to measure the flow rate of the water continuously . The 27 Figure 2 . 8 : Schematic of Generation 1 Water System The black lines around the water system an d the target indicate secondary containment. The dark er box around the target indicates a separate secondary containment outside of the main system. The dark blue components show important fittings including t - fittings that connect three tubing branches an d valves that can direct the flow of the water. F igure 2 . 9 : Images of Generation 1 Target and Water System at Beam End Station The target was housed in plastic secondary containment at the beam line end station (a) and all other water system components were enclosed in water - tight secondary containment adjacent to the beam line (b). 28 outputs of the conductivity, pH, and dissolved oxygen sensors as well as the flow meter were connected to a Mettler - Toledo M800 multi - parameter transmitter, which recorded read ings from each probe every 5 seconds. Two 3M Liqui - Cel membrane contactors (2.5 x 8 EXF, Membrane X40) were used to manipulate the dissolved gas content of the water. One of the membrane contactors (MC) was connected to a Marathon electric vacuum pump (M1 00GX, 1/6 HP) on the lumen side of the MC and was placed before the target to degas the water just before it entered the target volume. This reduced the number of possible air bubbles in the target to maximize heat transfer between the Ti64 disk and the wa ter. The second MC was connected to the hydrogen gas sensor on the lumen side and , allow ed H 2 to escape the water and be measured by the MIDAS gas sensor . Due to the operating requirements of the H 2 sensor, laboratory air flowing at 500 mL/min was used as flush gas to deliver H 2 to the sensor. This led to a re - equilibration of the water with atmospheric oxygen and carbon dioxide at the second MC , which in turn effected the readings of the dissolved oxygen sensor and the pH probe. The system water was also exposed to small pieces of platinized titanium mesh that were held in place with glass wool and a set of polypropylene fittings. This component, referred to as the degrader, was included as a catalyst to break down H 2 O 2 produced through radiolysis. This pr ocess was expected to proceed through the following reaction: [42] This reaction was utilized to break down the main long - lived corrosive radiolysis product, H 2 O 2 , to water and molecular oxygen. 29 C ation and anion exchange resin beds ( AG 50W - X8 and AG 1X8 , respectively; 20 - 50 mesh si ze; BioRad) depicted in Figure 2 .10 were made tubing and were connected to the rest of the syst em via polypropylene compression fittings and reducing unions on both sides of the tubing . About 10 - 15 mL of resin was us ed in each column and was held in place by glass wool plugs at the top and bottom of the resin beds . The larger tubing and low mesh siz e of the resin allowed a flow rate of 1 L/min through the columns with minimal back pressure. An additional set of resin beds constructed in the same way were used prior to the experiment to purify the water to a minimum conductivity of 250 nS/cm. These re sin beds were connected to the system by closing the valve to the ion exchange resin bed loop in the system and connectin g the tubing from either side of this loop to the resin beds used to purify the water. Figure 2 . 10 : Column Configuration for Aqueous Ion Collection The flow of water is indicated by the blue arrows from the top left of the figure, through both resin beds and ending at the bottom right of the figure. 30 The system was orga nized in this order for optimal sensor readings and conditions at each component. The first membrane contactor that served to degas the wate r was placed before the target and after re - equilibration with atmospheric air at the second membrane contactor and at the head space of the reservoir. The degrader was placed after the target so that the water coming from the target, with the highest conc entration of H 2 O 2 , was put in contact with the degrader material before going through the rest of the system. The me mbrane contactors, in particular, are sensitive to degradation from H 2 O 2 exposure. Two sensors in the system had competing flow rate require ments. The flow meter produced reliable readings at flow rates of about 1 L/min or higher . However, the pH probe had a pressure limit that required lower flow rates, significantly below 1 L/min in this system, to produce reliable readings. This conflict wa s solved by setting the flow rate at 1 L/min and placing the dissolved O 2 and the pH probe in parallel. This allowed for accurate flow meter readings while cutting the flow rate to the pH probe down to about 500 mL/min. Finally, the ion exchange resins wer e placed on a loop that could be bypassed using two valves to accomplish the two phases of the experiment: 1) the wa ter bypassed the resin beds during the irradiation to allow the accumulation of any ions resulting from corrosion and subsequent detection b y the conductivity probe and 2) after the irradiation, the valves were used to direct the flow of water over the res in beds to collect radionuclidic and stable ions from the water. 2 .3.1.2 Conclusions about the Generation 1 Water System The membrane con tactors contain a material, called a potting material, which maintains a water - tight seal between the lumen and the shell side and is sensitive to decomposition by H 2 O 2 . After the irradiation, water droplets were discovered in the tubing that connected the 31 lumen side of the first membrane contactor to the vacuum pump, indicating that the level of H 2 O 2 that built up in the system was sufficient to degrade the potting compound. It was concluded that other means of degassing the water should be used given the level of H 2 O 2 produced in the system and that the capacity for catalytic degradation of H 2 O 2 should be increased . An elemental analysis of the ions in the water system showed a large amount of potassium ions in the system. These ions mainly originated fro m the pH probe which uses a flow of counter ion from its electrolyte solution to measure the pH of the water. Th is source of ions interfered with measurements of corrosion with the conductivity probe. Exposure of the system water to the air for the H 2 meas urement at the second membrane contactor and through the vented headspace at the reservoir led to equilibration of the degassed system water with atmospheric O 2 and CO 2 . This interfered with both the measurement of dissolved O 2 and the pH level. These comp eting measurements demonstrated that more compatible sensors should be used and that when compatible sensors are unavailable, the most important measurements should be prioritized. These lessons were implemented to improve the design of the subsequent gene rations of water systems. 2 .3.2 Generation 2 and 3: 1 H and Low Intensity 48 Ca Irradiation Two experiments with a low intensity 48 Ca beam (Chapter 4 ) and a high intensity proton irradiation (Chapter 5 ) were performed with identical targets and very similar next - generation water systems. [36,40] Due to their similarities, these systems will be described collectively with the minor differences indicated. A schematic diagram of the isotope harvesting system used in low intensity 48 Ca irradiation is shown in Figure 2 .11 while that for the high intensity 32 proton irradiation is shown in Figure 2 .12. The system used for these experiments resembled the first - generation harvesting water system with adjustments for lessons learned from the first experiment. One impo r tant difference was the larger size of the system with 36 L of water flowing through many more feet of tubing. The water flow was supported by a centrifugal pump (Grundfos CRNE 3 - 6) running in constant pressure mode, as opposed to the pulse mode diaphragm pump used previously. The water was directed through multiple specialized loops, branching out from the 40 L water reservoir (McMaster Carr, portable stainless - steel dispensing tank). This allowed for more flexibility in the flow rates directed to each se c tion of the system. Flow to the target was adjusted by changing the pressure on the pump while each of the smaller loops had precision flow valves (Ham - let, 316 stainless steel) that could adjust their flow rates. The target loop had the largest water flow at 10 L/min, direct ing water through nylon - reinforced polypropylene tubing to a second - generation isotope harvesting target . The lines to the target contained non - spill coupling valves that could be disconnected and reconnected t o a series of large mixed bed resins (McMaster - Carr, Filter media PVC water deionizer) . The water was cycled rapidly through these resin beds to remove dissolved ions from the system water. This step was performed so that the final radionuclidic product wo uld contain the lowest pos sible level of stable ionic contaminants. Another section of the system , the hydrogen peroxide decomposition box, condition ed the water throughout each irradiation. The decomposition box received water at a rate of 300 mL/min thr ng. This loop was the most significant difference between the water systems used in the two experiments. During the low intensity 48 Ca 33 Figure 2 . 11 : Diagram of Isotope Harvesting System for Low Intensity 48 Ca Irradiation [36] Each loop in the system is designated by different dashed lines as shown in the figure legend. The circles at intersecting lines indicate a T - connection , the double triangles facing each other represent manual valves in the line , and the large circles represent sensors with the let ters specified in the legend. 34 Figure 2 . 12 : Diagram of Isotope Harvesting System for the Proton Irradiation [40] The solid outlines designate the different main subsystems: target loop, aqueous chemistry box, hydrogen peroxide decomposition box, and gas chemistry box. The solid circles at the intersection of more than two lines indicate a T - con nection, the double triangles with points facing each ot her indicate manual valves, and the large circles represent sensors with the letters specified in the legend. 35 experiment, the water was pushed through eight particulate filters ( Swagelok, SS - 6F - 60, 316 size ) each filled with 4 - 7 g stainless steel metal gauze (Alfa Aesar, 316 stainless steel, 100 or 200 mesh) in series. These filters were wrapped with heating tape strips ( HTS/Amptek, length 120 cm ), surrounded wit h insulation wool ( Knauf Insulation, high - temperature fiberglass ), and enclosed in a metal box. This construction maximized the catalytic decomposition capacity for H 2 O 2 by providing a large stainless - steel surface area and elevating the temperature of the water to about 45 ° C. Unfortunately, this experiment demonstrated that the stainless - steel surfaces gradually lost the catalytic decomposition ability, most likely due to the growth of a passivation layer. This H 2 O 2 decomposition unit was replaced with a catalytic converter ( Yonaka 2.5" ID Ultra High Flow Metal Core Race Universal Cat Catalytic Converter) for the high intensity proton irradiation . This proprietary unit contained noble metals ( i.e. , Pd and Pt), immobilized on a metallic honeycomb support, w hich can catalytically decompose hydrogen peroxide . [42] I n both experiments, this loop also contained a heat exchanger ( Brent I ndustries Titanium Pool Heat Exchanger SP - 55Kti - S ) with a counter - current of chilled water ( Thermo Scientific, Thermoflex 1400 W ) to maintain the system water temperature at ~25°C. This component was vital to remove heat deposited in the system water by th e high intensity beam and with the heating elements used in the decomposition box from the low intensity 48 Ca irradiation. Finally, there were separate aqueous and gaseous chemistry sec tions in the system to collect the radioactive products from the irradiation. Both loops were connected to the system The aqueous chemistry loop pass ed water over cation and anion exchange resins at 300 mL/min to remove any ionic species, stable or radioactive, from 36 Figure 2 . 13 : Drawing of Ion Exchange Resin Bed the water. The construction of the ion exchange resin beds is shown in Figure 2 .13. T he non - spill coupling valves ( Cole - Parmer, CPC non - spill quick - disconnect and coupling body ) on each side of the resin beds enabled easy and safe removal of the resin beds during an irradiation. This loop also includes a sample line with a manually o perated valve to allow for easy wi thdrawal of water samples from the system. The gaseous chemistry section of the system uses an acid and a base trap as well as a desiccant to remove species such as [ 13 N]NH 3 , [ 11 C]CO 2 , and water, respectively. Additional traps could be used to trap gas th rough cryogenics ( e.g. , stainless - steel traps cooled with liquid N 2 ) or with porous material ( e.g. , molecular sieves or activated charcoal). The gas dissolved in the water was carried by a purge gas that passed from the bottom of the reservoir water to the headspace at a rate of 200 mL/min. Gas released from the water or already in the headspace of the reservoir flow ed to the gaseous chemistry elements in the stream of purge gas. In the 48 Ca irradiation, a purge gas of ni trogen with 0.1% hydrogen was used a nd directed through the lumen side of the membrane contactor placed directly after the hydrogen decomposition box. In the proton irradiation, helium was used as the purge gas and the membrane contactor was 37 removed from t he system due to the sensitivity of the potting material to H 2 O 2 as previously discussed. With no membrane contactors in the system, all gas removal was performed with the purge gas through the reservoir. Throughout the system, the flow rate on each loop was monitored by a flow meter (Targe Peroxide Decomposition Box: Brooks Instrument, 1250 55 ; Aqueous Chemistry Box: Omega FTB323D Micro - Flo ; Gaseous Chemistry Box: Brooks In strument Sho - Rate Model 1250A ) . In a ddition, the conductivity sensors ( Mettler Toledo, UniCond Conductivity Sensor ) before and after the ion exchange resins provided a measurement of the overall conductivity in the system water and the effectiveness of the resins in removing the ionic specie s. Similarly, the dissolved oxygen sensors (Mettler Toledo, polarographic dissolved oxygen sensor) before and after the hydrogen decomposition box provide d information on the bulk dissolved oxygen in the system as well a s the breakdown of hydrogen peroxide (described above) , a reaction which produces molecular oxygen. A hydrogen sensor (H2Scan, HY - OPTIMA 710B) on the gas chemistry line monitor ed the production and evolution of molecular hydrogen. The temperature of the bu lk water was also monitored using te mperature sensors incorporated into the conductivity probes. A significant difference between the first and later generations of the water system was the lack of a pH probe and that the system was closed to atmospheric a ir. These changes allowed for the ef fective measurement of dissolved O 2 and ions due to radiolysis and corrosion, respectively. These sensors were monitored and controlled with a Programmable Logic Controller unit or PLC ( Automation Direct Productivity 2000 ) through the Productivity Suite ( v 3.3.0.17, Automation Direct ). 38 Several safety measures were put into place for chemical and radiological safety while operating the system. All components and tubing through which water or gas flowed were enclosed in secondary containment. The outlines arou nd different sections of the system shown in Figure s 2 .11 and 2 .12 indicate how the system was grouped into secondary containment units. The containment was achieved through a variety of boxes and carts Figure 2 . 14 : Images of the Isotope Harvesting System for the High Intensity Proton Irradiation Pictures showing the target at the end of the cyclotron line (a), the water system in secondary containment sitting behind a shielding wall (b), the inside of the aqueous chemistry box (c), and the inside of the hydrogen peroxide decomposition box (d). 39 enclosed with sheets of acrylic, and the tubing between each secondary containment unit was run through semi - rigid PVC tubing, shown in the pictures in Figure 2 .14. Additionally , the end of the gaseous chemistry line was equipped with air - tight gas collection bags to capture the stream of purge gas and any remaining radionuclidic or radiolytic gases not previously captured in the gas traps. Any remaining safety hazar ds ( i.e. , radioactive gases and radiolytically produced hydrogen gas) were collec ted rather than being vented into the experimental vault. The PLC unit which monitored and controlled the sensors in the system was also in system during the 48 Ca irradiation to facilitate a safety interlock on the cyclo tron RF ( i.e. , the frequency used to accelerate the beam) based on the water system readings. This system was designed to interrupt the irradiation of the isotope harvesting t arget if a large enough difference between a preset threshold and the measured pr essure or water flow was detected, among other interlocks. With this interlock, system failures would become readily apparent and the irradiation of the system would be automa tically stopped. 2 . 3.3 Generation 4: High Intensity 48 Ca Irradiations The final version of the water system used in this work was used in the two highest intensity 48 Ca irradiations discussed in Chapter 6 , 7 , and 8 and is shown schematically and with pi ctures in Figure 2 .15 and 2 .16. This water system was similar in design to that used in the high intensity proton irradiation with a few exceptions. The main difference in these systems was the separation of the different sections of the system. The isotop e harvesting target was p laced in the beam line at the beam blocker position after the first dipole magnet in the A1900 fragment separator (Figure 2 . 3 ), the aqueous and gaseous chemistry box were placed on the 40 Figure 2 . 15 : Schemat ic D iagram of Isotope Harvesting System for High Intensity 48 Ca Irradiation The three different locations that housed components of the system are delineated with dashed lines and indicated by the labels at the top of the drawing. The thick line around the components in the Transfer H all represents secondary containment around all these components. south wall of the N1 vault, and all other components ( e.g. , water reservoir, water pump, catalytic converter, PLC, etc.) were place d a t the east side of the tr ansfer hall where the A1900 fragment separator is housed. The main section of the system in the transfer hall was about 85 and 50 feet away from the target and the chemistry boxes, respectively. Placement of the target at the be am blocker position instead of an experimental end station allowed for a full power 80 pnA 48 Ca beam irradiation. While most of the components were held in the transfer hall, aqueous and gaseous chemistry boxes were placed in the N1 vault for easy access t o c ollection components duri ng the irradiation. Due to the distance between the components, longer lengths of tubing between the components and therefore a large water volume (approximately 50 L) were used in the system. 41 Figure 2 . 16 : Images of t he Isotope Harvesting System for Hig h Intensity 48 Ca Irradiations The main secondary containment unit in the transfer hall (a) and the aqueous and gaseous chemistry boxes in the N1 vault (b). Other changes to the s ys tem included Figure 2 .14) on the target loop with a much larger flow rate of 10 L/min through this unit as compared to the previously used 300 mL/min. This reduced the re sidence time of the water in the co n verter while increasing the number of passes of the water through the system in a given time period, potentially increasing the decomposition capacity of the unit. Another change in the main loop of the system was the us e of a new Teflon - coated brewing ke t tle as the water reservoir. The mixed bed resins used to purify the system water were also placed in a permanent, more easily accessible position after the aqueous chemistry box compared to the temporary insertion in the target loop used previously. Whi le this was more convenient placement, this configuration reduced the flow rate through the mixed bed resins, requiring more time to fully purify the system water. Another change in the aqueous chemistry box was 42 the additio n of lead shielding. A wall of le ad bricks was built on the bottom of the secondary containment in front of the cation exchange resin bed where the activity accumulated. A cylinder of shielding material was also constructed with a hole through the middle i n which the resin bed could be in se r ted. This provided shielding while users worked in and near the aqueous chemistry box and for the transportation and offline handling of the resin bed. A schematic drawing and pictures of the aqueous chemistry box interi or are given in Figure 2 .17. Fina ll y , additional interlocks between the isotope harvesting system and the NSCL beam control were integrated to protect equipment such as monitoring the beam line humidity in the case of a leak in the water - flowing target. Figure 2 . 17 : Aqueous Harvesting Loop for High Intensi ty 48 Ca Irradiation A schematic (a) and p hotograph (b) of the isotope harvesting box, lead shielding around a cation exchange resin bed (c), and a cation exchange resin bed (d). In the s chema tic on the left, the column is indicated by the grey rectangle, valve s are depicted as blue double triangles facing each other, T - i s drawn as light blue lines. 43 2 .4 Conclusion Efforts in the isotope harvesting group have led to the development of a reliable and flexible water system for the production and recovery of radionuclides produced by the interaction of high - energy heavy - ion beams with flowing - water target. S e v eral experiments have been performed to provide information on the detailed water chemistry during an irradiation and to design an isotope harvesting system for safe and convenient u se for research purposes. Chapter 3 presents the results of an experimen t with a low intensity 40 Ca irradiation that used the first - generation flowing water target and isotope harvesting system to test the durability of Ti64 as the target material. This te st was followed by a higher intensity proton irradiation discussed in Ch a p ter 4 that tested the durability of the second - generation isotope harvesting target and the third - generation water system at NSCL - scale beam power depositions that more closely resem bled the conditions anticipated at FRIB. The target and system durabili t y demonstrated at these beam power levels allowed for full power irradiations of the second - generation water target ( i.e. , 0.5 kW power with 80 pnA 140 MeV/nucleon 48 Ca beam) at the N SCL (Chapters 6 and 7 ). The first 48 Ca irradiation of a prototype flowin g - water target used the second - generation flowing water target and started with a low intensity 48 Ca irradiation ( i.e. , < 1 pnA) discussed in Chapter 4 and lead to higher intensity 48 C a irradiations ( i.e. , 1 - 80 pnA) presented in Chapters 6 and 7 . The secon d - generation water system was used in an experimental vault for the first 48 Ca irradiation, while the fourth - generation water system was used with the isotope harvesting target in the A1900 beam blocker position for the final higher irradiations. This seri e s of experiments provided valuable 44 information for the development of a system and techniques that will facilitate isotope harvesting at the higher beam powers available at FRIB. 45 Chapter 3 : L ow Intensity 40 Ca Beam Ti64 Durability T e st 7 A flowing - water target was irradiated with a 140 MeV/u, 8 nA 40 Ca 20+ beam to test the feasibility of isotope harvesting at the upcoming Facility for Rare Isotope Beams. Among other radionuclides, 2.6(2)E - 6 48 Cr and 5.6(5)E - 6 28 Mg nuclei were formed t hrough fusion evaporation and fr a gmentation reactions, respectively, for e very impingent 40 Ca and were collected on an ion exchange resin bed . Radiolysis - induced molecular hydrogen evolved from the target at an initial rate of 0.91(9) H 2 molecules per 100 eV of beam energy deposited. No r adiation - accelerated corrosion of the tar get material was observed through radioactive and stable element tracing and surface analysis . 3 .1 Introduction Previous isotope harvesting experiments at the NSCL have involved th e implantation of secondary radio a ctive beams into a static volume of water [32 34] . These experiments demonstrated success in recovering the radioactive products from an aqueous medium and subsequent chemical purification and use in applications such as nuclear medicine. With these e ncouraging results, subsequent isotope harvesting expe riments were designed to simulate more closely the harvesting conditions expected at FRIB in the future. A new water system was developed (see Chapter 2 Section 2 .2.1 and 2 .3.1) to include a flowing - wat er target with a 7 The material presented in this chapter draws heavily from the published paper E. Paige Abel, Hannah Clause, and Greg Severin, Radiolysis and radionuclide p roduction in a flow ing - water target during a fast 40 Ca 20+ irradiation, Appl. R adiat. Isotopes. 2020 , 158 , 109049. 46 beam window of Ti64 alloy, the materi al that will be used in the FRIB isotope harvesting beam dump. [39] Attached to this target was a system that monitored and co nditioned the water as well as collected radioactive p roducts from the flowing water. Additionally, realistic isotope harvesting conditions at the NSCL or in the future, at FRIB, will involve a much higher intensity beam of primary ( i.e. , stable) particles . With a higher intensity beam, the implanted beam par ticles interact to a greater extent with the water through both nuclear and chemical reactions. The extent and effect of these reactions should be studied through a series of isotope harvesting experime nts with increasing beam intensity. The nuclear reacti ons that occur in the isotope harvesting target allow for production of radionuclides from accelerated primary beams ( e.g. , the production of 47 Ca from a 48 Ca primary beam). The important nuclear intera ctions in this context are fragmentation (FR) reaction s at high energies , and fusion - evaporation (FE) reactions at low energies . FR reactions result from high energy interactions, where beam particles collide with target nuclei and break apart, forming fra gments of the original nucleus. As the energetic beam loses energy in the target, FE reactions begin to occur, primarily at the end of the particle tracks. It is important to know the extent to which these reactions occur to be able to predict the activity of a particular radionuclide of interest and the over all level of activity produced for experimental and safety planning, respectively. These reaction rates can be measured directly through experiments or predicted through s imulation codes like PACE4, Lis Fus, and LISE++. [29,43,44] Additionally, it is important to test if the radionuclid es produced in a water medium react chemically as expe cted for planning methods or collection, purification, and use in applications. 47 In addition to these nuclear reactions, chemical reactions also occur as a result of implanting a fast, heavy ion beam int o a water target. As the beam particles pass through t he target, the water molecules are ionized and undergo dissociation reactions in a process called radiolysis. This produces a variety of ionic, radical, and molecular species such as H + , OH - , OH, H , H 2 , O 2 , and H 2 O 2 . While some of these products quickly recombine especially when produced by high linear energy transfer (LET) particles, the molecular species in particular can be long - lived. [45 48] Conventionally, the production rate of these molecular species have been quantified as the G - value, which is defined as the numbe r of molecules formed per 100 eV of energy deposited. In the present isotope harvesting systems, the measured value gives an understanding of the survival rate of the longer - lived species since they are measured after recombination of radiolysis products o ccurs in the target volume. This rate is of interest sin ce many of the radiolysis products such as the radical products and H 2 O 2 form a corrosive environment in the water and gases such as O 2 and H 2 can lead to a relatively high pressure and potentially ex plosive condition. Measurements of the long - lived radiol ysis in the isotope harvesting system with increasingly high intensity irradiations can demonstrate the extent to which these products buildup, recombine, or decompose as experiments come closer to FR IB conditions. To test the nuclear and chemical reaction s in a water target, an experiment was carried out at the NSCL with a stable primary beam impinged on the first - generation flowing - water target. The production rate of several radionuclidic products w ere measured and compared to predicted rates with PACE4, Lisfus, and LISE++. The activity collected on ion exchange resin beds were measured to gain an understanding of the collection efficiency with this method and the 48 chemical behavior of radionuclidic p roducts formed in the water target. Additionally, H 2 gas was removed from the system and measured as an indication of the extent of radiolysis in the target. The potential corrosive effects of radiolysis were monitored by looking for corrosion of the water side of the Ti64 beam window in the target through an a nalysis of the stable elements and radionuclides collected on the ion exchange columns. Altogether, this experiment provided a small - scale test of isotope harvesting at the NSCL and FRIB. 3 .2 Materia ls and Methods 3 .2.1 Isotope Harvesting System Irradiation A 40 Ca +20 beam was accelerated to an energy of 140 MeV/nucleon by the Coupled Cyclotron Facility at the NSCL. The beam was initially delivered to the target at the end of an experimental beam line at an intensity of 0.11(1) particle nanoamperes ( pnA ) or 2.1(3) elec trical nanoamperes ( enA) for approximately 17 minutes. After this initial test, the intensity was increased to 0.42(4) pnA or 8.4(9) enA for about 3.5 hours. These beam current readings we re obtained by using frequent unsuppressed current readings from the target and intermittent current readings from an upstream calibrated Faraday cup. The unsuppressed measurements ( i.e. , proportional but not equivalent to the true value) were scaled using the calibrated beam current measurements. The current measurements from the unsuppressed target, calibrated Faraday cup, and scaled target readings are shown in Figure 3 .1. The scaling of the unsuppressed target current readings was performed using three averages for each of the two beam current intensities: the average Faraday cup measurements, the average target measurements above a certain threshold, and the average of all the target measurements. These thresholds were set at 2 enA 49 for the lower intens ity beam setting and 5 enA for the higher intensity beam setting to represent both the main band of measurements as well as the lower fluctuations in the measurements that can be observed in Figure 3 .1. Since the Faraday cup measurements were intermittent, they did not reflect the fluctuations in the target current reading s. To make the Faraday cup measurements more representative of these lower fluctuations these readings were scaled by the ratio of the average of all target measurements divided by the ave rage of the target measurements above the thresholds. These ratios w ere found to be 0.89 and 0.97 for the lower and higher intensity beam settings, demonstrating that the lower fluctuations in the beam current were small relative to the beam current measur ements throughout the irradiation. These ratios were used to find a scaled average of the calibrated beam current values for both beam intensity settings. Additionally, the ratio between the scaled average Faraday cup readings and the average target readin gs (1.71 and 1.17 for the lower and higher Figure 3 . 1 : Scaled Beam Current During Experiment The beam current recorded every 5 seconds by the unsuppressed target (teal) and intermittently recorded by an up stream calibrated faraday cup (purple) are plotted. The Faraday cup readings were used to calibrate the unsuppressed target readings, resulting in the scaled beam current readings in the plot (green). 50 beam intensities, respectively) were used to scale the frequent target readings as plotted in Figure 3 .1. With ratios greater than 1 for both beam intensity settings, the scaled target readings are higher than the unsuppressed target current measurements. The target used for this experiment was made of a flo w ing - water filled polyethylene body with a disk of grade 5 titanium alloy (Ti64, 6% aluminum and 4% vanadium by mass, balance titanium) beam window (Figure 2 .1, Section 2 .2). This outer shell of polyethylene and Ti64 was needed to contain the flowing - water target medium. The beam passed through several layers of material that were used to separate the water from the beam line vacuum and contain the water , depicted in Figure 3 .2, before passing through the Ti64 disk and finally entering the water volume insi d e the target. The energy of the beam as it traveled thro ugh these layers was estimated by using stopping powers obtained from SRIM and is shown in Figure 3 .2 . [49] The surface of the disk facing the water was divided i n to two halves, with one Figure 3 . 2 : Depiction of Layers Traversed by 40 Ca 20+ Beam The 140 MeV/nucleon 40 Ca beam (black dotted line) passed through two layers of air, a Zr foil, and a Ti foil before passin g through the Ti64 target disk and into the interior water layer of the target. The energy of the beam particles before entering the target set up, before t he target front face, and before entering the water in the target is given in the figure. 51 half mil l e d smooth and the other half left as received after additive manufacturing through 3D printing. During the experiment, the beam was directed to enter the t arget through the half of the Ti64 disk that was not milled. A flowing - water system containing compo n e nts such as sensors, membrane contactors, ion exchange resin beds, a water reservoir, and a water pump was attached to this target (see Figure 2 .8 and Sec tion 2 .3.1.1). Approximately 6 L of ultra - pure water was circulated through the system. Prior to irr a d iation , the water flow was directed through a set of cation and anion exchange resin beds (AG 50W - X8 and AG 1X8, respectively) to reduce the conductivity of the water to about 250 nS/cm. The baseline conductivity increased to approximately 800 nS/cm afte r removal of the columns, presumably due to re - equilibration with atmospheric CO 2 across the post - target membrane contactor. The pH of the water was also co nsistent with atmospheric CO 2 equilibration, ranging from 5.1 to 5.8 throughout the experiment. Afte r irradiation, the water flow was redirected sequentially through a cation exchange (AG 50W - X8, 20 - 50 mesh) and an anion exchange resin bed (AG 1X8, 20 - 50 m esh) at a flow rate of 1 L/min to remove ions from the water. 3 .2.2 Radionuclide Quantification an d Comparison to Production Estimates Following the ion exchange process, components of the water system were removed and analyzed using a Canberra BEGe Gamm a - ray Detector (BE2020). An energy and efficiency calibration were performed using a 152 Eu point sou r c e at 50 cm from the face of the detector. Gamma spectra in this work were primarily taken at 50 cm from the detector. In cases where counting rates required a different distance, the detector efficiency was corrected mathematically rather tha n using a re c a libration. 52 The physical components analyzed with gamma spectroscopy included the water in the reservoir; the target cell; the hydrogen peroxide degrader; and the cation and anion exchange columns. The spectra allowed for identification, quan tification, a n d localization of radionuclides on these components . The shorter - lived species were quantified using spectra obtained within a few hours after the irradiation. To reduce background from the shorter - lived nuclides, the longer - lived species wer e quantified f rom spectra obtained approximately 2 days after the irradiation. The characteristic gamma - ray energies and the branching ratios used to identify and quantify the radionuclides are given in Table A1 Appendix A [14,15,28,50 56] . Contributions to the 511 keV gamma - ray peak from radionuclides such as 34m Cl, 43 Sc, 44 Sc, 48 V, 48 Cr, a n d 52 Mn were calculated based on the activities found on each component using the characteristic gamma rays listed in Table A1 . Since the remaining counts demonstrate a decay half - life close to that of 18 F ( i.e. , a measured half - life of 116.9 min from four d ata points vs. the accepted half - life for 18 F of 109.77 min), they were used to make an estate of the amount of 18 F produced. Due to interferences such as a high Compton background and their relatively short half - lives ( 43 Sc t 1/2 = 3.891 h, 44g Sc t 1/2 = 3 .97 h) , the activity of 43 Sc and 44g Sc i n some system components could not be measured. In particular, 43 Sc was not observed in the water, 44g Sc was not observed on the cation excha nge resin bed, and neither scandium isotope was observed on the hydrogen p e roxide degrader. T he values were estimated u sing the observed scandium isotope activities with the assumption that the distribution of all scandium isotopes on the system components is the same due to their identical chemistry. The longer - lived 44m Sc (t 1 / 2 = 58.61 h) was quantified two days after the irradiation with low - background spectra. The 53 distribution of 44m Sc on the anion exchange resin bed , the cation exchange resin bed , the hydrogen peroxide degrader, and in the system water was then used to find c orresponding activities of 43 Sc and 44g Sc in these locations. In practice, t he known activities for 43 Sc and 44g Sc on these components were scaled by the proportion of 44m Sc on each system component to find estimates for the unknown activities of 43 Sc an d 44g Sc. The calculated values for these scandium isotopes are labeled in Table 3 .1 . For comparison to the observed activity, cross - sections obtained from PACE4 and LisFus through the program LISE++ and stopping powers obtained from SRIM were used to estim a t e the expected FE reaction production rates for 43 Sc, 44,44m Sc, 48 V, 48 Cr, and 52 Mn. These radionuclides were assumed to be produced solely through FE reactions between 40 Ca and 16 O. T he cross section s from either PACE4 or LisFus were used in Equation 3 . 1 to find a production rate, P , in terms of particles produced per incoming particl e for each of the radionuclides observed: ( 3 . 1) where represents the cross section of a fusion - e v a poration product at a given energy , is the stopping power of the beam in the target materia l at a given energy, and are the density and molar mass of the target material, and is Avogadro's number. The integral was approximate d as a sum over energy steps from to , the initial and final energy of an interacting particle in the target, respectively. The production rate in Equation 3. 1 is the number of nuclei produced per incoming 40 Ca 20+ and was converted to partic l e s produced per second by multiplying by the incomi ng beam intensity . The cross sections from PACE4 and LisFus are given as a function of beam energy in Tables B1 - 10 in Appendix B . 54 FR production rates were calculate d with LISE++ using a primary beam of 40 C a 20+ at an energy of 137.59 MeV/nucleon ( i.e. , the resulting energy after passing through the first four layers of materials shown in Figure 3 .2) and an intensity of either 0.11 pnA or 0.42 pnA. Layers representing the target face , the water volume, and a d etector material were specified in this order in the program to find a production rate in the unit particles produced per second during the experiment. The production values found for all methods previously described were used wi th the typical activation e quation to find the final predicted activities. 3 .2.3 Radiolysis Measurement The production of H 2 gas was measured as an indicator of radiolysis in the water target during the irradiation. While the H + and O 2 levels in the water were measured with a pH m e t er and a dissolved oxygen sensor, these levels were influence by the equilibration of the water with atmospheric air. In particular, the pH was decreased by the equilibration with atmospheric CO 2 in the air and the dissolved O 2 l evel was increased by the e quilibration with atmospheric O 2 . Therefore, the measured levels of these radiolysis products were not a direct indicator of radiolysis with the isotope harvesting system used in this experiment. Previous experiments have reporte d G - values for ion beams a t different average linear energy transfer (LET) values [47,57 59] . The average LET values were found using Equation 3 .2: ( 3 .2) where is the initial energy and is the stopping power of the ion beam. [59] In practice, this equation can be approximated as a sum from the initial energy to the end of the beam track. A logarithmic trend has been found be tween the average LET values vs. the measured G - values 55 for different ion beams. Examples of such relationships are shown in Figure 3 .3 for 60 Co gamma rays, a 1 MeV 1 H beam, a 5 MeV 4 He beam, and fission fragments [47,57 59] . The average LET calculated value for the 40 Ca beam as it passed through the water layer of the target was used with this relationship to predict a G - value of 1.57 H 2 molec ules/100 eV of deposited beam energy. Throughout the experi ment, the H 2 gas production was measured with a Midas gas detector at the second membrane contactor in the system. This detector used a sweep gas of air moving at 500 mL/min through the detector and measured H 2 in this diluted stream in ppm. The measured c oncentration of H 2 gas was converted to the number of H 2 gas molecules produced per second during the irradiation, by estimating the dilution factor from the sweep gas using the ideal gas law. To find an experimental G - value for H 2 , this production rate wa s Figure 3 . 3 : Average LET vs. Experimental G - Values The logarithmic trend between the average LET value and experimentally determined G - values f or various ionizing radiation sources from the literature [47,57 59] . This trend was used to predict a G - value for the 124.3 MeV/nucleon 40 iation. 56 combined with the measured beam current and the energy of the beam as it entered the water as estimated with SRIM. The irradiation was separated into a different segment each time the calibrated Faraday cup was inserted to measure the beam current. At these points, the beam no longer reached th e target and no H 2 gas was produced through radiolysis. As expected, the level of H 2 gas measured at these short intervals rapidly dropped to zero. The G - value corresponding to the average beam current and H 2 gas production rate during each irradiation per iod are presented and discussed in Section 3 .3.3. 3 .2.4 Corrosion Assessment 3 .2.4.1 Ion Exchange Resin Processing After measuring the gamma spectra of the columns to quantify and locate the radionuclides th at were produced and collected on the ion exchange resins, both the cation and anion columns were processed to remove any stable ions collected from the water. First, the water in the tubing above the resins was drained and saved. The cation column was the n treated with solution of increasing acid molarity in the following order: about 11 mL of 0.6 M HNO 3 , 12 mL of 2 M HNO 3 , 12 mL 3 M HNO 3 , 12 mL 4 M HCl, and 12 mL 8 M HCl. After each addition, an equivalent volume of solvent was pulled through the column a nd removed in small fractions. Gamma spectra of each group of eluents were taken after removing 12 mL between each addition to ensure ions were being removed. In a similar way, the an ion column was treated with the following solutions : 12 mL water, 12 mL 2 M HCl, 12 mL 4 M HCl, and 12 mL 8 M HCl. Between each addition, an equivalent amount of liquid was removed from the columns in small fractions and analyzed with gamma spectroscopy as for the cation resin samples. This 57 acid elution gradient was performed on the ion exchange resins to remove as many ionic species as possible from the columns [60,61] . All fractions from the cation and separately the anion resin were later combined and analyzed by inductively coupled plasma - optical emission spectroscopy (ICP - OES) with an Agilent TCP700 spectrometer. This analysi s was performed to quantify any stable elements that accumulated in the system water during the irradiation. First , a semi - quantitative ICP - OES method was used to identify the elements present as well as a rough order of magnitude estimate for the concentr ation of elements present in the cation and anion resin eluates . The following elements were found to be present a t a concentration of approximately 0.1 ppm or higher: B, Na, Mg, Al, Si, P, S, K, Ca, Fe, Ni, Zn. Further q uantification of these elements was performed by preparing standards containing these elements in 2% nitric acid (v/v). The samples from the cation a nd anion resin were each diluted to a total acid concentration of about 0.7 M HCl and were run with the prepared standards and a blank of 0.7 M HCl on ICP - OES. The measured concentrations are presented and discussed in Section 3 .3.4.1. The wavelengths used for quantification of each element are given in Table C1 in Appendix C . 3 .2.4.2 Surface Assessment Images of the Ti64 target window were tak en with a Perkin Elmer IVIS Lumina LT In Vivo Imaging System (CLS136331) to visualize the location of the beam str ike. First, a picture was taken of the disk with the camera in the IVIS imaging system. With the disk in the same place and orientation, a thi n piece of scintillating plastic was placed over the disk and a second picture was taken to localize the area of t he disk that was activated by the beam strike. These 58 images were then overlaid to verify the location of and attribute any visual differences to the beam strike. The resulting image is presented and discussed in Section 3 .3.4.2 3 .2.4.3 Corrosion Rate Estim ation An analysis of the corrosion rate of the target window was also performed to understand the effects of radiolysis and the energy deposit ion of the beam on the target window. This analysis used 47 Sc produced in the Ti64 beam window as a radiotracer to find an upper limit for the corrosion rate of the window material. T he methods for estimating fusion evaporation cross sections described pre viously indicated that the cross section for 47 Sc production in the water was very low. Through gamma spectroscopy , 47 Sc was found as a product in the target window. Therefore, any 47 Sc found in the water or accumulated on other system components would ind icate target degradation. After the irradiation, no detectable amount of 47 Sc was measured in the water or on any of the other system components. Based on this measurement, the limit of detection of 47 Sc with HPGe measurements of these components was used as the upper limit for the degradation of the target material. The limit of detection was approximated as the uncertainty in the sum of background counts where the 159 keV peak from 47 Sc would have been detected in each gamma spectrum considered. Specifi cally, t he sum of counts from 158.8 to 160.3 keV was found for each spectrum and the error in this region was taken as the square root of the sum of counts. This uncertainty was then set equal to the number of counts that could have resulted from 47 Sc in t he system water but were not observed due to statistical counting errors. Spectra in which 44m Sc was detected were used in this analysis: the anion exchange resin bed, the cation exch ange resin bed, the hydrogen peroxide degrader, and the water in the rese rvoir. It was 59 assumed that 47 Sc would have accumulated on these components as well since all scandium isotopes behave the same chemically. As three of these components ( i.e. , the cati on exchange column, the anion exchange column, and the hydrogen peroxide degraded) contained glass wool, it appears that they filtered out neutral but aqueous forms of scandium such as HScO 2 and the rest remained in the water [62] . It was also assumed that the proportions of 44m Sc found in each area of the system was equivalent to the distribution of 47 Sc. Using the count limit and the proportions of 44m Sc found on each component, values were calculated for the activity of 47 Sc that could have also accumulated on each of the other three components ( e.g. , the count limit found for the anion resin was used to find the 47 Sc activity that could have accumulated on the cation resin, the hydrogen peroxide degrader, and the water reservoir) . The total amount of 47 Sc that could have b een in the system water without being detected was the sum of activity found in this way for all four parts of the system where Sc isotopes were detected. The lowest act ivity sum was on 3 .4 below). Th is total activity of 47 Sc was used to model the target degradation as a time dependent process. The 47 Sc activity was built up in the target using smal l time steps and by assuming a production rate that was proportional to the beam intens ity: ( 3 .3) where P was a constant production rate that produced a final estimated activity in the window equal to that measured value. This production rate was multiplied by a factor involving the integrated beam intensity in each time step where is the difference between and . The sum of activity in the window at each time step was found by adding th e decay corrected 60 activity from the previous st ep whe re is the difference between and . It was also assumed that 47 Sc would enter the water from the target as a constant percent of the 47 Sc present in the window per unit time: ( 3 .4) where F is the fraction that degraded per unit time, L is the total activity limit determined as described previ ously, and the denominator is the sum of activity across all the time steps. In this way, a percent degradation per hour and over the course of the e xperim ent was found. It is important to note that this method limits the degradation rate of only the beam strike area on the target window since this is the only part of the window in which 47 Sc was produced. Therefore, a limit on the degraded mass at the beam spot was calculated using the relative area of the beam spot to the whole disk and the mass of the wh ole disk. 3 .3 Results and Discussion 3 .3.1 Isotope Harvesting System Irradiation During the irradia tion of the harvesting system, the temperature w as mea sured by the dissolved oxygen sensor, the conductivity probe, and the pH meter. Since the temperature s recorded by these three sensors were found to be comparable, only the temperature data from the dissolved oxygen sensor will be discussed here. The tempe rature of the bulk water, as measured by the dissolved oxygen sensor over the course of the experimen t, is shown in Figure 3 .4. A temperature change of 0.93 ° C was observed over the course of the irradiation, resulting from the deposition of approxim ately 28 kJ in the water and 3 kJ in the target window [49] . Based on the energy and current of the beam as well as the volume of water running through the system, a maximum temperature increase of 1. 10 - 1. 21 ° C was ant icipat ed. The upper and 61 lower limits of this range were found with the ass umption that all of the heat or none of the heat deposited in the target face , respectively, was removed by the flowing water. Since the estimated temperature increase of the water d ue to beam energy deposited directly in the water was approximately 18% hi gher than the measured temperature increase, p assive cooling of the system water most likely occurred to a small extent to produce this difference. The temperature of the target wind ow dur ing this irradiation was not a concern since only about 9.5% of the total energy of the beam is deposited in the Ti64 beam window , the beam current was low throughout this short irradiation, and cooling occurred as the 2 4.3 - 2 5 .3 ° C water flow ed by th e targ et window at 1 L/min. Assuming this cooling by the water occurred ac Ti64 window were estimated by evaluating the energy deposition and the heat transfer between these materials for short time s teps throughout the irradiation time. This procedure resulted in an estimated temperature of 25.5 ° C for the water and 25.7 ° C in the Ti64 window ( i.e., +1.2 ° C and +1.4 ° C temperature change, respectively). Based on these estimates, it is likely tha t the water provided sufficient cooling of the Ti64 beam window and that p assive cooling of the water occurred in this experiment. The conductivity of the water was measured throughout the experiment to provide a qualitative measurement of the level of di ssolve d ions in the water. These conductivity measurements taken before an d during the irradiation are shown in Figure 3 .4. Initially , a saturation behavior is observed as the water came into equilibrium with the carbon dioxide in the air. This provided a baseli ne of about 800 nS/cm for the conductivity of the water before 62 Figure 3 . 4 : Measurements of the T emperature and Co nductivity of the W ater during the I rradiation The temperature and conductivity of the bu lk water are shown as a function of the time elapsed since the water was purified using ion exch ange resins. Each irradiation period is indicated by a different color: before irradiation (red), 2.1 enA irradiation (gold), 8.4 enA irradiation (blue), and af ter irradiation (green). the irradiation. During both irradiation periods, the conductivity inc reased linear ly . However, the magnitude of the slope increased with the higher intensity beam current, indicating a higher production of ionic species in the wa ter as the intensity increased. Additionally, the 63 sudden decreases shown in Figure 3 .4 correspon d to times when the Faraday cup was inserted to measure the beam current. Less stable ionic species such as radicals with short lifetimes were no longer p roduce d through radiolysis at the target, corresponding to a decreased conductivity. After this initia l decrease when the F araday cup was inserted, the presence of relatively stable ionic species in the water was shown by a constant conductivity reading. A t the end of the irradiation, ion exchange columns removed both stable and radioactive ions from the w ater, lowering the conductivity to about 100 nS/cm. This low conductivity value demonstrates the effective, nearly complete removal of ions from the irrad iated water. At some point during the approximately 4 - hour irradiation, a small leak developed at the flowing - water target. This led to less than 100 mL of water leaking from a gap between the target body and the back plate of the target. This volume, at less t han 2% of the total water used, falls within the precision with which the total water volume was known. Therefore, the effects of this leak on quantifications of the stable and radioactive ions in the system was deemed negligible. Based on an inspec tion o f the target following the irradiation, two important observations were made : (1) The O - ring whi ch provided a seal between the target body and the back plate of the target was out of place and had been noticeably stretched. This indicated that pressu re ins ide the target put stress on the seal at the back plate . (2) The screws that held the back plate on the target body developed s ignificant oxidation . This most likely occurred as the beam entered humid air before reaching the target. In these conditio ns, a high energy particle beam can create a corrosive atmosphere of ozone and nitric acid in the air. This environment potentially compromised one or more of the screws securing the back plate of the target. It is 64 also possible that the temperature of the bulk water as recorded by the sensors in the system did not provide an accurate understanding of the local temperature in the target volume. Additionally, an increased pressure in the target could have led to flexibility in the polyethylene target shell m ateria l, making the seal at the back of the target less secure. Any one of these possibilities could h ave led to the minor leak that was observed at the target. This target set up w as not used further, as a target shell made completely of the Ti64 alloy wa s used for the next several experiments. 3 .3.2 Radionuclide Quantification and Comparison to Production Estimates The gamma - ray spectra shown in Figure 3 .5 demonstrate the activity collected on the cation resin, the anion resin, the hydrogen peroxide degr ader, and the reservoir water. Table 3 . 1 provides the quantification by gamma - ray spectroscopy of each radionuclide on these components . Additionally, the production rate of these radi onuclides in atoms produced per incident 40 Ca ion was calculated based o n the measured total activities and the beam current during the irradiation and are given in T able 3 . 1. The location of the radionuclides collected in the system matches the expected chemical speciation [ 62] . For instance, sodium, magnesium, and manganese were collected on the cation exchange res in and chlorine, vanadium, and chromium were found on the anion exchange resin. For the most part, fluorine was also collected on the anion exchange resin. The re mainin g activities from beryllium and scandium isotopes were found on a few system components; this was expected as they are known to form neutral species in an aqueous solution that is slightly acidic and oxidizing. The efficiency of the system for ion ex tracti on was demonstrated by the very low level of activity found in the water after it was ci rculated over 65 the ion exchange resins. Most radionuclides were not observed in the water above the background in the measurement. Even 18 F, the main radionuclide found in the reservoir water, was collected with an efficiency of 94(1) %. Table 3 . 1 : Quantification of and Production Rate Estimate for Radionuclides Produced in Water Target The quantification of each radion uclide s was decay corrected to the end of bombardment (EOB) and was determined wi th HPGe spectroscopy. The activities denoted with an asterisk (*) are calculated activities based on the distribution of 44m Sc on the system components. Radionuclide Activity Meas u red (kBq) Production Rate (atoms produced per incident 40 Ca atom) Cation Resin Anion Resin Degrader Water Total 7 Be 13.9(4) 3.3(2) - - 17.2(4) 3.2(3) X 10 - 3 18 F 39(9) 1.194(5) X 10 3 - 80(4) 1.31(1) X 10 3 6.6(7) X 10 - 4 24 Na 125(7) - - 6(1) 130(7 ) 3.0 ( 3) X 10 - 4 28 Mg 1.7(1) - - - 1.7(1) 5.6(7) X 10 - 6 34m Cl - 4.5(9) X 10 3 - - 4.5(9) X 10 3 1.7(4) X 10 - 3 43 Sc 11(1) 23.5(9) 10.0(6)* 11(1)* 56(2) 4.3(5) X 10 - 5 44g Sc 4(1)* 6(3) 3(1)* 2.4(7) 16(3) 1.2(3) X 10 - 5 44m Sc 1.15(3) 1.59(4) 0.81(1) 0.83(9) 4 .4(1) 3.8(4) X 10 - 5 48 V - 0.54(3) - - 0.54(3) 3.0(4) X 10 - 5 48 Cr - 0.79(6) - - 0.79(6) 2.7(4) X 10 - 6 52 Mn 0.68(4) - - - 0.68(4) 1.3(6) X 10 - 6 66 Figure 3 . 5 : Example Gamma - Ray Spectra of System Components Gamm a - ray spectr a with their live counting time given in parenthesis: ( a ) cation resin on the day of the irradiation (520 s), ( b ) the cation resin two days after the irradiation (2959 s), ( c ) the anion resin on the day of the irradiation (517 s), ( d ) the anion exch ange resin two days after the irradiation (3957 s), ( e ) the hydrogen peroxide degrader two days after the irradiation (429 s), ( f ) the reservoir water on the day of the irradiation (1275 s), ( g ) and the target window two days after the irradiati on (13 97 s) . All unlabeled peaks in the spectra are at energies corresponding to known background gamma rays. 67 Production estimates for the FR and FE reaction products were performed using LISE++, LisFus, and PACE4. The estimates as well as the measured ac tiviti es ar e shown in Table 3 .2 for the FR reactions and Table 3 . 3 for the FE reactions. These values were calculated for the average beam current values over the two different intensity sections of the irradiation: 0.11 pnA for 17 minutes followed immedia tely b y 0.4 2 pnA for 3.5 hours. Since the p roduction distribution between the metastable and the ground state of radionuclides such as 44 Sc and 34 Cl are not given by codes, the following distributions were used for these radionuclides . The projected isomer ic rat ios f rom both PACE4 and LisFus were found to be closer to the observed ratio when the production proportion was 3:1 44m Sc: 44g Sc. Attributing 100% of the production rate for 34 Cl calculated with LISE++ to producing 34m Cl achieves an activity of 3.2 X 10 3 kB q which is within two sigma of 4.5(9) X 10 3 kBq the observed activity for 34m Cl. Activities corresponding to th ese distribution s for 44 Sc and 34 Cl are given in Table 3 . 3. Table 3 . 2 : Activity M easured and P redi cted for R adionuclides P roduced through F ragmentation Reactions. The asterisk (*) denotes that the ratio of produced 34g Cl to 34m Cl is not known so 100% of the production rate found with LISE++ was attributed to making 34m Cl as this assumed producti on dis tribution provides a predicted activity closest to the measured activity. Radionuclide Activity (kBq) Measured Predicted with LISE++ 7 Be 17.2(4) 1.4 18 F 1.32(1) X 10 3 1.6 X 10 3 24 Na 130(7) 246 28 Mg 1.7(1) 14.3 34m Cl 4.5(9) X 10 3 3.2 X 10 3* 68 Tab l e 3 . 3 : Activity M easured and P redicted for R adionuclides P roduced through F usion E vaporation Reactions . The asterisks (*) denote that information about the percent of 44 Sc produced in the metastable vs. the groun d stat e is unknown. Instead, percentages were found that most closely matched the produced 44m Sc to 44g Sc rati o. The predicted activities shown in the table for both PACE4 and LisFus were found, assuming 25% of the 44 Sc produced is 44g Sc and 75% is 44 m Sc. Radio nuclide Activity (kBq) Measured Predicted with PACE4 Predicted with L isF us 43 Sc 56(2) 62.5 148 44g Sc 16(3) 22* 14* 44m Sc 4.4(1) 6.03* 3.77* 48 V 0.54(3) 1.30 0.39 48 Cr 0.79(6) 5.18 19.0 52 Mn 0.68(4) 2.00 0.68 The methods used here for esti mating production in the water - filled target vary in the accuracy of their predictions . While the estimates for 18 F, 24 Na, and 34m Cl are inaccurate by only a factor of four or less, the predictions for 7 Be and 28 Mg differ by about an order of magnitud e fro m obse rvation. For the fusion evaporation estimates, the values are mostly within a fa ctor of 4 of the measured values. The exception is 48 Cr with values differing by a factor of 7 for PACE4 and 24 for LisFus. For the level of activity produced in thi s sho rt, lo w - intensity test, the differences in predicted and measured values did not prese nt a safety hazard or an experimental hurdle. However, as the irradiation length and intensity increase, it will become more important to correctly estimate producti on ra tes. T his will allow adequate radiological controls, and guide harvesting efforts to p roducts that will be present in significant enough quantities for use in off - line experiments. 69 3 .3. 3 Radiolysis Measurement To observe the behavior of the hydrogen meas uremen ts within each irradiation period, the measured hydrogen concentration was overl aid on the beam current in Figure 3 .6. As in Figure 3 .1, the scaled beam current readings from the unsuppressed target are shown as the large number of points in gre en wh ile th e calibrated beam current readings from the Faraday cup are shown as the smaller number of points in purple. The hydrogen gas measurements are given on the secondary y - axis and are represented in pink. The graph shows that shortly after a Farad ay cu p meas urement was taken, during which the irradiation of the water was interrupted, th e hydrogen concentration drops significantly. This occurs as the hydrogen production through radiolysis is interrupted during these disruptions in the irradiation of the water. Additionally, the decrease in H 2 concentration is not instantaneous after the F araday cup insertion as it took a short amount of time for the accumulated hydrogen gas to escape the system as the hydrogen - rich water passed by the second membrane cont actor in the system. A qualitative trend in the hydrogen concentration is shown in Fi gure 3 .6 for the higher intensity irradiation setting. During the first period at this intensity setting, the highest hydrogen concentration occurred at the beginnin g of the ir radiation period. Within this period, however, the measured hydrogen concentrati on decreases. The second higher - intensity irradiation period starts at a lower concentration than the first and then exhibits a similar trend of decreasing concentra tion measur ements through the irradiation period. The hydrogen concentration appears to rea ch a lower limit during the last three higher - intensity irradiation periods. 70 An average effective G - value was found for each of the six irradiation periods (Table 3 .4). With t his value, the hydrogen production during the irradiation periods can be quantit atively compared, as the effective G - value factors in the average beam current and the average dilution factor for the hydrogen concentration measurement for each ir radia tion p eriod. The effective G - value for the radiolytic production of hydrogen gas demo nstrates the decrease in hydrogen production that is visually apparent in Figure 3 .6. However, with the normalization of the beam current in the effective G - value me asure ment, it is shown that the highest relative hydrogen production occurred in the first irradiation period at the lower - intensity beam setting. The largest difference between the effective G - value of two beam intensities occurred between the first two i rradi ation periods with smaller differences thereafter. While the trend demonstrated in Fi gure 3 .3 could be improved with more data points and a more comprehensive fitting function, it demonstrates a clear relationship between experimentally measured G - va lues for H 2 and the average LET of a particle. Using this trend, a G - value for the producti on of hydrogen gas was predicted at almost double the maximum rate observed experimentally in this irradiation. One practical reason for the low observed value could be d ue to the permeability of most plastic materials to H 2 gas. The isotope harvesting sys tem used here produced H 2 gas at the target, sent the hydrogen - rich water through a section of polyethylene tubing, and measured the H 2 gas that escaped from the wat er th rough a membrane contactor. This set up potentially allowed for some escape of the pro duced H 2 gas. A possible explanation for the decrease in H 2 concentration measured as the irradiation progressed is the occurrence of recombination events between ra dioly sis pr oducts. While many 71 radiolysis products recombine rapidly within the particle tra ck, some molecular species such as H 2 , O 2 , and H 2 O 2 remain for are longer - lived in the water. Buildup of these products, particularly H 2 O 2 in the water could have af fecte d reco mbination within the particle tracks, resulting in a suppressed experimental val ue for hydrogen production resulting in the measurement of an - value. [40] Most likely, the effe ctive G - values presented here are suppressed values compared to the true production rate. However, these measurements provide an understanding of the overall level of H 2 that results from production and recombination reactions in an isotope harvesting irra diatio n. The suppression observed here may also indicate that lower level s of radiolysis products than predicted will be observed in future irradiations and that radiolysis product levels will be more manageable than anticipated. More measurem ents of a wid er ran ge Table 3 . 4 : Avera ge Hydrogen Gas Production and Experimental G - v alue for Six Irradiation Periods A 5% uncertainty was used for the measured H 2 varia nce in repeatability for a measurement of less than ± 5%. Irradiation peri od Average Beam Current (enA) Measured H 2 Concentration (ppm) Experimental G - value (H 2 molecules/100 eV) 1 2.0 130(7) 0.86(4) 2 8.69 459(23) 0.70(4) 3 8.69 438(22) 0.67(3) 4 8.3 3 91(20) 0.62(3) 5 8.43 392(20) 0.62(3) 6 8.52 392(20) 0.61(3) 72 Figure 3 . 6 : Hydrogen Gas Production and Beam Current The hydrogen gas measured in ppm and the beam current in en A are shown as a function of irradiat ion ti me. Although the beam current remains relatively steady for about four hours of the irradiation, the measured hydrogen gas production decreases throughout this time period. of radiolysis products have been performed (not as a part of the present work) to better understand the dynamics of radiolysis products in the isotope harvesting system, particularly as irradiations use increased beam intensities to more closely match those expected at FRIB. 3 .3.4 Corrosion Assessment 3 .3.4.1 Ion E xchange Resi n Proc essing The vast majority of the ions in the system water were collected on the cation and anion exchange resins used following the irradiation. This is demonstrated by the very low conductivity and the small activity of 18 F in the bulk wa ter after it was p assed o ver these resins. Therefore, the concentrated samples collected from the resins should provide a quantitative 73 understanding of the ions present in the system water at the end of the irradiation. In the ion exchange column eluate, t he following eleme nts wer e identified and quantified: B, Na, Mg, Al, Si, P, S, K, Ca, Fe, Ni, Zn. Table 2.5 provides information about the amount of each element that was found on either the cation or the anion exchange resin. Table 3 . 5 : Quantification of S table E lements E luted from I on E xchange R esins after I rradiation, C ollection, and E lution The values in this table are measured using ICP - OES. The upper limits provided for vanadium are estimates of the limit of detectio n from this measurement. Element Mass Eluted (mg) Cation Resin Anion Resin Total Recovered from Columns B - 0.50(2) 0.50(2) Na 1.45(4) 0.22(2) 1.67(5) Mg 0.17(2) 0.39(2) 0.57(3) Al 0.07(3) 0.66(3) 0.72(4) Si 0.10(9) 3.28(7) 3.4(1) P 0.1 3(2) - 0.13( 2) S 0.034(3) 0.229(3) 0.264(4) K 67(2) - 67(2) Ca 0.52(5) 0.29(3) 0.81(6) V < 0.006 <0.009 <0.014 Fe 0.022(1) 0.029(1) 0.051(1) Ni - 0.035(2) 0.035(2) Zn 0.064(6) 0.021(1) 0.085(6) 74 When removing ions in the water with the resins, the water circul ated o ver first the cation exchange resin and then the anion exchange resin. The cation exchange resin had a H + counter ion while the anion exchange resin had an OH - counter ion. This arrange ment was used so that overall , the two counter ions w ould recombi ne to form water molecules instead of adding additional ions to the system ( e.g. , Na + and Cl - ) and increasing the conductivity of the system water . However, the chemical state of the anion resin with an OH - counter ion may have had the unexpect ed effect of preci pitating a fraction of the Ca 2+ and the Mg 2+ in the water as it passed over the resin. While the presence of Na+ on the anion exchange resin is also unexpected, this may have re sulted from residual traces of the NaOH rinse used prior to t he experimen t to c onvert the anion resin to the OH - form. The potassium level found on the cation resin most likely resulted from the pH probe in the system. This probe uses a KCl electrolyte so lution and the pH measurement depends on this solution passin g through a semipe rmeable membrane into the test solution. This means that the electrolyte was passing through the probe into the system water for several hours as the beam was tuned, during the irradiation, and while the ion exchange resins removed the ac cumulated io ns fro m the water after the irradiation. Finally, the level of aluminum found on the anion exchange column was a concern at first, as no aluminum containing parts were expected to be in contact with the water except the Ti64 beam window disk t hat containe d 6% A l by mass . The source of the aluminum is now thought to be the screws that held the backplate of the target in place. Although these were black oxide steel screws, dissolved alu minum was detected in both dilute HCl and H 2 O 2 that had been in contact w ith ne w screws. These screws were not intended to be exposed to the water 75 but when a small leak resulted from over - pressurizing the system, the backplate loosened. This allowed the sc rews to be exposed to irradiated water, leading to a small am ount of alum inum i n the system water. The hypothesis that the aluminum did not originate in the target window is supported by the observation that no vanadium was found in the eluate from either column. Since 48 V, a product from the irradiation, was found on the anio n exch ange resin and nowhere else in the system, it is to be expected that stable vanadium would have been mainly collected on the anion exchange column as well. Therefore, vanadium w ould have been detected at least in part in the samples elute d from the a nion e xchange resin. A limit of detection estimate is included in Table 3 .5 demonstrating that the possible level of undetected vanadium in the eluate was very low, not exceeding abou t 0.015 mg total. An estimate of the amount of vanadium expec ted in the e luate if the aluminum resulted from target degradation can be made based on two assumptions: 1) vanadium and aluminum had the same removal efficiency from the columns and 2) the amoun t of vanadium corroded from the window is related to the amou nt of alumin um cor roded by the mass proportions of each element in the window material (6% aluminum and 4% vanadium in Ti64). Based on these reasonable assumptions, 0.48 mg of vanadium should hav e been detected in the column eluate. This is a significant a mount of van adium, well above the limit of detection using ICP - OES. Therefore, it is very unlikely that this amount of aluminum resulted from target degradation. Except for the potassium level, the amount of each stable ion removed from the system is fair ly low. Howe ver, t his analysis of the stable ion contaminants in the system provides information on the identity and relative level of competitive ions for future separations , 76 radiolabeling exper iments , and other chemical processes or measurements with har vested radio nuclid es. 3 .3.4.2 Surface Assessment The two images of the irradiated Ti64 disk described above are shown in Figure 3 .7. The most notable result is that the beam activation spot on the disk image in Figure 3 .7b appears at the same spot on the disk where a visu al difference can be observed in the photograph shown in Figure 3 .7a. This visual difference between the irradiat ed and non - irradiated areas of the Ti64 disk could also be observed on the disk with the human eye. Since this difference is so readily a pparen t, this indicates significant changes to the water - facing side of the disk at the beam strike. This change likely resulted from the ionizing beam of 40 Ca 20+ passing through the Ti64 material and from radiolysis products in the water mediu m in contact with the disk at the beam strike. These conditions could induce changes in the internal structure of the alloy ( e.g. , atom displacements) and changes to the passivation layer of the alloy in contact with the water Figure 3 . 7 : Optical Image of Irradiated Ti64 Disk The optical image from the Ti64 without (a) and with (b) a scintillating plastic sheet over the Ti64 disk. Both images were taken with the disk in the same placement in the instrument, allowing fo r the verification of the beam strike location in the left p icture of the disk. The luminescence scale to the far right is an uncalibrated scale that indicates relative intensity of activity in the beam spot. 77 interior of the target, respectiv ely. While th ese e f fects may lead to corrosion of the Ti64 target materi al, it is also possible that they could stimulate a hardening effect of the material that will make it more corrosion resistant. Further studies are necessary to resolve these questio ns since the exten t of material erosion observed in this experiment was s mall. 3 .3.4.3 Corrosion Rate Estimation Using gamma spectroscopy, a 47 Sc activity of 16.4(12) kBq was measured in the Ti 64 beam window. No equivalent activity of 47 Sc was found o n the components. A lim i t of detection estimate was used to set an upper limit on the activity of 47 Sc that could have entered the water from the window. The gamma spectrum of the hydrogen peroxide degrader and the relative proportions of 44m Sc found on va rious system compo nents provided an upper limit of 60(10) Bq of 47 Sc could have degraded into the water ( i.e. , less than 0.4% of the 47 Sc measured in the Ti64 beam window) . The amount of 47 Sc in the window and the water was assumed to build up in the wate r ove r time. The p roduc t ion in the window follow ed the beam intensity, while the degradation from the window was assumed to be a constant percent of the activity in the window. The 21(3 ) % per hour o 0 .7(1 ) % over the c o urse of the 3.5 h experiment . The difference between the 0. 4 % upper limit of 47 Sc degradation determined with gamma spectroscopy and the 0.7 % upper limit on the beam spot degradation can be explained by the time dependent production of 4 7 Sc in the be am wi n dow. Although the degradation of the window was assumed to be constant, the 47 Sc accumulated in the window throughout the irradiation. Only if the final measured 47 Sc activity in the Ti beam window was present from the beginning of the i rradiation wo uld t h ese two upper limits match. 78 The percent degradation can also be c onverted to the amount of mass degraded from the beam spot. Using the mass of the beam spot only, the amount of mass degraded could have 18(4 ) mg, which was 0.02 % of the tota l tar g et disk mass . It should be emphasized that this value is an upper limit of the degradation of the Ti disk at the beam spot, meaning the actu al degraded mass was equal to or smaller than the limit presented here . 3 .4 General Discussion Th is isotope ha rvest i ng experiment demonstrated the ability to use a flowing water - filled target to produce and harvest radionuclides from a high - energy heavy - io n primary beam. Several radionuclides were identified and quantified as reaction products from ei ther fragment ation or fusion evaporation reactions. The collection efficiency of the ion exchange resins in the system was quite high as only scandium isotopes and a small amount of 18 F was observed through gamma spectroscopy in the reservoir water followi ng collection on i o n exchange resins. The conductivity of the water also dropped to 100 nS/cm after using the resins, indicating the removal of most stable ion s as well. This test of the initial prototype system show ed the feasibility of producing and usin g radionuclid es ge n erated through isotope harvesting. During this experiment, the isotope harvesting system was tested under the effects of an irradiation. A m inor leak developed, showing that the target system used in this experiment is not suitable for h andling this level of irradiation effects such as an increased temperature and pressure. This experiment was intended to test the durability of the target wind ow material, Ti64, under bombardment by beam particles and the harsh radiolysis effects in the in terior water. Ther e fore, a new target system was designed in which the entire target shell will be 79 made of Ti64. This target was designed to handle high pressu re, elevated temperatures, and high flow rates , making it unlikely that leaks w ould develop at th e target in f uture experiments. Additionally, this experiment used a low beam intensity over a short irradiation period, resulting in only a minor increase in temperature. As future irradiations increase in intensity, changes were made to the system includ ing active co oling with a water chiller, larger water volumes, and higher flow rates. Furthermore, for the 400 kW beam power anticipated at FRIB, the beam will irradiate a circulating target, which will spread the deposited energy over a large area on the target face [39] but a rotating target is beyond the scope of the present work . Each of these adjustments will allow for more efficient cooling of the target face than with the meth ods used in t his e x periment ( e.g., cooling the target window with 6 L of water flowing at 1 L/min and passive cooling of the water in the reservoir). The production and effects of radiolysis products were also explored during this irradiation. The hydrogen gas produced thro u gh radiolysis was measured at the two different beam intensities and was compared to theoretical production estimates. The amount of hydrogen gas measured was lower than the predicted level, indicating that radiolysis products in this sy stem may reco mbine to a larger extent than previously thought or might be lost through diff usion through plastics . Additionally, the value continued to decrease slowly as the irradiation progressed, suggesting that the build - up of radiolysis products in th e water led t o a l o wer amount of H 2 being produced or escaping from the target system . Alth ough this experiment did not provide a G - value measurement for pure water radiolysis, the observed hydrogen production was indicative of what can be expected in futu re isotope ha rvest i ng experiments at NSCL and FRIB. Relative measurements of radiolytic pro ducts in future 80 experiments can provide a benchmark that will indicate changes in water chemistry. This will be important as upgrades are made to the target and wate r purificatio n sys t em and higher beam intensities are used. The effects of radiolysis on the system components, especially the target window, were explored using an analysis of the stable and radioactive ions in the system water as a result of corrosion . Since the le vels o f stable ions detected in the system were relatively low, there was no i ndication of wide - spread system degradation. Although aluminum was detected in the system, no corresponding vanadium was observed with ICP - OES. This strongly suggest s the aluminu m did not originate in the target disk but from a different component in the s ystem. Additionally , 47 Sc was not found in the system water or on any components in the system even through it was produced in the target disk. T ogether, these obser vations stron gly i n dicate that no target disk degradation was observed in this preliminary experiment. 3 .5 Conclusion The results of this experiment demonstrate d that isotope harvesting with a stable ion beam and a flowing - water target is a viable way to produce radio nucli d es for research purposes, give a measurement for the production of H 2 as a radiolysis product, and provide evidence that Ti64 may be able to withstand the effects of an irradiation and the resulting radiolysis product s. The comparison of predicted to meas u red activities of the radionuclides produced in this irradiation demonstrated the importance of measuring the production rates for accurate estimates and experimental planning. As the measurement of H 2 gas production in this experiment w as far lower than p redicted using literature G - values, further experimental measurements of the radiolysis species produced in the isotope harvesting system are also 81 necessary to understand the extent of the effect of radiolysis in isot ope harvesting exper iments. The r esult s of this experiment and improvements to the isotope harvesting target and water system (see Chapter 2 Section 2 .2.2 and Section 2 .3.2) facilitated the next stage of isotope harvesting experiments with low intensity 48 Ca and 78 Kr beams at the NSCL. 82 Chapter 4 : Low Intensity 48 Ca Irradiation for the Production, Collection, and Purification of 47 Ca 8 An experiment was performed at the National Superconducting Cyclotron Laboratory using a 140 MeV/nucleon 48 Ca beam and a flowing - wat er ta r get to produce 47 Ca for the first time with this production route. A production rate of 0.020 ± 0.004 47 Ca nuclei per incoming beam particle was measured. An isotope harvesting system attached to the target was used to collect radioacti ve cationic p roduc t s, including 47 Ca, from the water on a cation exchange resin. The 47 Ca collected was purified using three separation m ethods optimized for this work: 1) DGA extraction chromatography resin with HNO 3 and HCl, 2) AG MP - 50 cation exchange r esin with an incre a sing concentration gradient of HCl, and 3) AG MP - 50 cation exchange resin with a methanolic HCl gradient. These method 47 Ca with 100% radionuclidic purity within the limits of detection for HPGe spectroscopy . ICP - O ES was used to identify low levels of stable ions in the water used in the isotope harvesting system during the irradi ation and in the final purified solution of 47 Ca. For the first time, this experiment demonstrated the feasibility of t he production , col l ection, and purification of 47 Ca through isotope harvesting for the generation of 47 Sc for nuclear medicine applicatio ns. 8 The material presented in this chapter draws heavily from the published paper E. Paige Abel, Katharina Domnanich, Hannah Clause, Colton Kalm an, Wes Walker, and Greg Severin, Production, Collection, and Purification of 47 Ca through Isotope Harvesting at the NSCL. ACS Omega , 2020 , 5 , 27864. 83 4 .1 Introduction After testing the first - generation isotope harvesting flowing - water target and system (described i n Chapter 3 ), the e quipment was upgraded to a target shell made completely of Ti64 and a larger water system ( see Chapter 2 Section 2 .2.2 and 2 .3.2). T his new system was used for isotope harvesting with a high - energy 48 Ca beam to produce 47 Ca for the first time with th is pr o duction route. In a wide range of oxidation potentials and pH values, calcium is expected to be in the Ca 2+ form in wa ter. This simple chemistry should allow for the 47 Ca produced in this harvesting experiment to be easily removed from t he water and proce s sed in the laboratory with high yields. The ease of working with calcium is a major advantage to harvesting 47 Ca for use in a 47 Ca/ 47 Sc radionuclide generator. Other el ements that are easily hydrolyzed in near neutral pH conditions, such as 48 V and 8 8 Zr, h ave also been produced through isotope harvesting. [63] These radionuclides, with more complicated chemistries, have proven less easily collected and chemically modified offline. 84 4 .2 Materia ls an d Methods 4 .2.1 Materials 4 .2.1.1 Reagents Before the irradiation, the water in the isotope harvesting system was purified using mixed bed resins (McMaster - Carr, Filter media PVC water deionizer), resulting in a conductivity level of 250 nS/cm. Chemic al pr o cessing of the products was performed with the following reagents: hydrochloric acid (VWR Chemicals, ACS grade, 36.5 - 38%), nitric acid (VWR Chemicals, ACS grade, 68 - 70%), methanol (Macron Fine Chemicals, anhydrous, ACS grade), and MilliQ water (Therm o Sci e ntific MicroPure Ultrapure Water System, 1 4 .2.1.2 Extraction Chromatography and Ion Exchange Resins - tetra - n - octyldiglycolamide, normal resin, particle size 50 - into a column and sequentially pre - rinsed with 20 mL of 5 M HCl, 5 M HNO 3 , and MilliQ water before use in the separations. Two cation exchange resins (AG50W - X8, mesh size 20 - 50, BioRad and AG MP - 50, 100 - 200 mesh size, BioRad) and an anion exchange resin (AG8X1, mesh size 20 - 50, BioRad) wer e prepared in large quantities by rinsing with the following solutions: 50 mL of 2 M HCL, 50 85 mL of 4 M HCL, 50 mL of 6 M HCL, and 100 mL MilliQ water. The rinsing steps described here were performed twice to remove ionic impurities, especially metallic imp urities, from the resins before loading the resins into the columns . 4 .2.1.3 Column Construction The columns used for collection and separation were made of rigid polycarbonate - to - connect fitt ings with a PBT (polybutylene terephthalate) body and stainless steel tube gripping clamps on both ends of the each end of the column to stabilize the resin. These columns were inserted in the water system isotope harvesting system (Figure 4 .1). Figure 4 . 1 : Harvesting System Over view 86 4 .2.1.4 Instruments Identification of stable ions was performed with an Agilent Inductively Coupled Plasma - Optical Emission Spectrometer ( TCP700.) Identification and quantification of radionuclides were performed with an HPGe Canberra BEGe Gamma - ray Detector (BE2020). Energy and efficiency calibrations of this detector were previously performed with a 152 Eu point source 50 cm from the detector face. Analysis of spectra was performed with Genie 2000 software (Mirion Technologies). Even t hough non - standard sample geometries were used for gamma spectrometry, no correction factors or additional errors were considered in quantifying radionuclides. Measur ements were performed that demonstrate d about a 10% difference in the quantification of these nuclid es between a point source geometry and a water sample at 25 or 50 cm from the detector face and less than 10% for that between a point source geometry and i ons adsorbed on a resin bed also at 25 or 50 cm from the detector face . Absolute quantification base d on the water samples was only performed to determine the total activity and production rate for 47 Ca, where the 19 - 20% error in the branching ratios far o utweighs the uncertainty from the geometry [8] . When quantification was performed for radionuclides on the cation exchange resin beds, separation columns, or small volumes in falcon tubes at 25 cm from the detector face, no additional uncertainty for the geometry was used as it is a small correction. These measurements taken for the sep aration methods were also used to calculate the percent activity eluted , so the absolute quantificat ion was not necessary. 87 4 .2.2 48 Ca Irradiation A 140 MeV/nucleon 48 Ca 20+ beam was used to irradiate a flowing - water target over 8.5 hours. Beam current measurements were automatically recorded every second on average from the unsuppressed target ( i.e. , the recorded values were proportional but not equivalent to the true beam c urrent. ) Intermittently, an intercepting F araday cup was inserted into the beam to get an absolute measurement of the current. The measurements with the Faraday cup were used to calibra te the concurrent unsuppressed readings from the target. Figure 4 .2 sho ws a linear relationship between the unsuppressed target current readings and the calibrated readings from the Faraday cup. This relationship was used to scale the beam current readings on the target measured on average every second throughout the experime nt. After a short tuning period at an intensity of 0.26 pnA, the beam intensity was increased and maintained at an average of 0.92 pnA for approximately 5.1 hours. Figure 4 . 2 : Calibration of Target Current Readings The current recorded on a calibrated faraday cup and the unsuppressed target are given in electrical nanoamps (enA). 88 The irradiation was paused occasionally for samples of the system water to be withdrawn from the system and ion exchange r esin beds to be removed and inserted . According to the timeline in Figure 4 .3 , three 1 - liter water samples were taken during the irradiation, and a fourth sample was taken after the end of the irradiation (named Water Sample s 1 - 4 for reference , respectivel y). After Water Sample 3 was withdrawn, a cation exchange resin bed (AG 50W - X8, mesh size 20 - 50, H + form, 1.5 g, 8.9 cm x 0.6 cm ID) was inserted in the system with a water flow of approximate ly 180 mL/min over the resin bed. This cation exchange resin bed was then replaced with fresh resin beds after two more irradiation periods (Resin 1, 2, and 3, respectively). Resin 3 was left in the system for two hours after the irradiation ended with the flow rate increased to 500 mL/min for the second hour to increa se the collection rate. Two more cation resin beds (Resin 4 and 5, respectively) were put in the system in parallel as well as an anion exchange resin bed in series the day after the irradiati on . F igure 4 . 3 : Timeline of Irradiation The beam current (blue), water sample collection (green), and resin bed changes (orange) are shown as a function of time. The beam current shown are the calibrated va lues of the current on the target . 89 A final 1 L water sample was taken at the end of the collection effort (Water Sample 5). Altogether, activity was collected in five 1 - liter water samples and five cation exchange resin beds that were used for further me a surements and experiments. 4 .2. 3 Production of 47 Ca The activity of 47 Ca found in the water samples and on the cation exchange resin beds was measured with the HPGe detector. Analysis of these spectra was performed with Genie software to detect and integ r ate peaks, implement baseline co rrections, and calculate efficiencies at each peak energy. This allowed for the detection and quantification of radionuclides based on their characteristic gamma ray energy emissions. P hotop eaks from three characteristic ga m ma rays were used to quantify th e activity of 47 Ca: 489.2 keV at 5.9 ± 1.2%, 807.9 keV at 5.9 ± 1.2%, and 1297.1 keV at 67 ± 13%. [51] The total produced activity of 47 Ca was estimated as the sum of activity found in each water sample, on each of the collection ion exchange resin beds, and in the water after col lection with the resin beds. The activity remaining in the system water was approximated by measuring the activity in a water sample taken when the last column was removed from the system and scaling up this activity to acc o unt for the total remaining wate r volume of 33 ± 4 L. This large uncertainty in the water volume mainly resulted from difficulties in draining the water from all components and tubing in the system. Due to this 11% uncertainty in the total water volume of the system and 19 - 20% uncertainties in the branch ing ratios, the estimated activity has a large associated uncertainty. For future measurements of this production rate, other indirect methods of measurement will be used to determine the system water volum e more accurately and these branching ratios will be remeasured to reduce the associated errors. 90 The total produced 47 Ca activity and the recorded beam intensity throughout the irradiation were used to find the production rate, of 47 Ca in the flowing - water target in terms of particles produced per in coming beam particle with Equation 4 .1: ( 4 .1) where is the number of produced nuclei , is the beam current during the i th irradiation interval from to , is the decay constant of the produced radionuclide, and is the time between the i th irradiation interval and the end of the irradiation. This segmented product ion equation accounts for fluctuations in beam intensity during the irradiation. The production rate of 47 Ca in a water target has also been predicted using two simulation codes that predict the production rate of radionuclides in nuclear reactions: PHITS (Particle and Heavy Ion Transport code System) and LISE++. [29,64] For both estimates, a model of the target was used in the program to account for a 500 m layer of Ti alloy followed by a water layer . Additionally, both fragmentation reactions (at higher particle energies) and fusion evaporation reactions (at lower particle energies) with 16 O and 1 H were included in the production rate estimates. Finally, the measured production rate for 47 Ca in this experiment was used to predict the production at higher beam intensities available at the NSCL and FRIB. Since the 48 Ca primary beam expected to be available at FRIB will be accelerated to a higher energy than the currently available 48 Ca beam at the NSCL, LISE++ was used to find the relationship between the production of 47 Ca at 140 and 189 MeV/nucleon. This comparison was used to qualify the accuracy of this extrapolated activity prediction for FRIB. The setti ngs used 91 in LISE++ to find a predicted product ion rate for 47 Ca and to compare the production of 47 Ca at NSCL and FRIB energies are given in Appendix B. 4 .2.4 Collection and Sample Processing T he 47 Ca activity collected on five cation exchang e resin beds was removed with 50 - 70 mL of 3 M H NO 3 per resin bed with a flow rate of 1.8 - 2.0 mL/min. This eluate from each resin bed contained a mixture of cationic radionuclides and was separated into several fractions for Figure 4 . 4 : Collection, Elu tion, and Preparation of Cationic Radionuclides Sample processing was performed in this manner except for samples used with separation method 1, which were used directly after elution from the collecti on resin. 92 use in three separation methods. Method 1 us ed the eluate directly as the load solution. Fractions used for methods 2 and 3 were evaporated to dryness on a rotary evaporator and reconstituted in 0.1 M HCl and 0.5 M HCl/90% methanol, respectively. See Figure 4 .4 for a schematic overview o f the sample processing. 4 .2.5 Purification of 47 Ca The following three methods were developed and optimized in this work to separate Ca from Na, Mg, K, Sc, and Fe. These elements were present as 24 Na, 27,28 Mg, 42,43,44,45 K, 44m,47,48 Sc, and stable Fe ( i.e. , natural) d uring the experiment. After each separation, the columns were rinsed with water for storage. 4 .2.5.1 Se paration Method 1: DGA resin with HCl and HNO 3 A column of 1.07 g of DGA resin (dry packed, 7.2 cm x 0.6 cm ID) was preconditioned with 20 mL of 3 M HNO 3 and a 10 - 20 mL sample of the collection column eluate in 3 M HNO 3 was loaded at a flow rate of 1.8 mL/ min. These conditions were chosen to elute many of the co - produced radionuclides such as 24 Na, 28 Mg, and 43,44 K as well as stable Fe while retaining 47 C a on the resin. The flow rate was decreased to 1.3 mL/min and an additional 10 - 15 mL of 3 M HNO 3 was used to rinse the column and ensure that all Na, Mg, and K isotopes were entirely removed. Elution of 47 Ca was carried out with approximately 20 mL of 3 M HCl to selectively remove 47 Ca and leave any Sc isotopes adsorbed on the resin. 4 .2.5.2 Separat ion Method 2: AG MP - 50 with HCl A column of 1. 5 g AG MP - 50 resin (slurry packed, 7.5 cm x 0.6 cm ID) was pre - conditioned with 0.1 M HCl. Evaporated fractions f r om the collection columns were re - 93 dissolved in 20 mL of 0.1 M HCl. This concentration was used s ince Ca has a large distribution coefficient with AG MP - 50 at this condition, allowing for the creation of a narrow 47 Ca band during the loading step. This sol u tion was loaded on the column and was followed by rinse steps of 10 mL of 0.1 M and 23 - 25 mL of 2 M HCl. Rinsing with these HCl concentrations allowed for the elution of co - produced radionuclides such as 7 Be, 24 Na, 28 Mg, and 43,44 K and any stable Fe. The n , 10 mL of 5 M HCl was used to elute 47 Ca while retaining the Sc isotopes on the column. This se paration was carried out with a flow rate of 0.8 mL/min throughout. 4 .2.5.3 Separation Method 3: AG MP - 50 with Methanolic HCl 1. 0 g of AG MP - 50 resin (slurry p a cked, 5 cm x 0.6 cm ID) was pre - conditioned with 0.5 M HCl in 90% methanol. Evaporated fractions from the collection columns were reconstituted in 20 mL of 0.5 M HCl in 90% methanol and the solution was loaded on the column. Rinse solutions of 0.5 M HCl/9 0 % methanol (20 mL, 1 mL/min), 2 M HCl/60% methanol ( 50 mL , 0.75 mL/min), and 2 M HCl/30% methano l ( 35 mL , 0.75 mL/min) were used in succession . The load and rinse solution of 0.5 M HCl/90% methanol was chosen due to the high distribution coefficient of Ca and the low distribution coefficient of Fe under these conditions. Therefore, any stable Fe pres ent could be eluted in this step while forming a narrow 47 Ca band on the column during loading. The two intermediate rinse steps were used to elute coproduced r adionuclides such as 7 Be, 24 Na, 28 Mg, and 43,44 K. Specifically, the 2 M HCl/30% methanol rinse s tep was added during optimization of this separation to allow for complete separation of the K isotopes from 47 Ca. Elution of 47 Ca from the column was perform e d with 15 mL of 4 M HCl at a flow rate of 1.25 mL/min. The high distribution coefficient of Sc f or all rinse media used in this separation ensured that any Sc isotope s remain on the column. 94 4 .2.5.4 Separation Yield and Radionuclidic Purity For each separ a tion performed, the separation yield was calculated as the activity of 47 Ca in the purified fractions divided by the total activity loaded on the column for the separation. The radionuclidic purity was found as the activity of 47 Ca in the purified fractio n s divided by the total activity in these fractions. Both the separation y ield and radionuclidic purity of the purified 47 Ca were found using the rate at which characteristic gamma rays for each radionuclide were observed by the gamma detector to avoid usi n g the reported branching ratios and their large associated errors for 47 C a. This was made possible as all spectra were taken at 25 cm from the detector face and the yield and purity were calculated in terms of ratios, avoiding the need for absolute activi t ies. Therefore, the only error considered in the gamma - ray rates was the statistical counting errors. Since the yield and radionuclidic purity of the 47 Ca for each of the separation methods was quite high ( i.e. , 100% yield and radiopurity), the limit of d etection (LOD) for radionuclides that would affect these values was also found. For the yield, the LOD for 47 Ca was the sum of the LOD for gamma spectra of (1) the fraction taken just before 47 Ca was observed in the eluate, (2) any fractions taken after 4 7 Ca was no longer observed in the eluate, and (3) the column after all fra ctions were taken for a separation. These were the samples that were most likely to contain 47 Ca that was at or below the LOD . The largest influence on the radio nuclidic purity was a n y activity below the LOD for 43 K, since this was the radionuclide of the highest activity that eluted close to 47 Ca. Tailing elution of this radionuclide could have occurred into the 47 Ca elution peak, affecting the radionuclidic purity. Therefore, the L O D of 43 K was found in each fraction that was part of the total yield of 4 7 Ca. 95 In determining the LOD, only the highest intensity gamma - ray energy was considered for each radionuclide ( i.e. , 1297 keV for 47 Ca and 372 for 43 K). The LOD was taken as the unc e rtainty in the counts over a range of 3.5 keV for 43 K and 5 keV for 47 Ca centered at their most intense characteristic gamma - ray energy. The total LOD for each of these values was the sum of the limit in each spectrum considered. The limit was then conver t ed to a percentage in terms of the total 47 Ca activity ( i.e. , the sum of eluted 47 Ca activity in each separation). This LOD was smaller than the error associated with the separation yield and the radionuclidic purity in all cases. Any activity of 45 Ca tha t was present in the purified 47 Ca fractions was not measurab le due to the absence of reasonably intense gamma - rays from this radionuclide ( i.e. , the only gamma ray has an energy of 12.47 keV at an intensity of 3E - 6%). While this calcium isotope, if presen t , would follow 47 Ca through all separation methods, it would also follow 47 Ca through the previously published pseudo generator. [10] Ad d itionally, 45 Ca decays slowly (t 1/2 = 162.61 d) to stable 45 S c, which would not affect the radionuclidic purity and only minorly affect the specific activity of the final radiolabeling solution. Since 45 Ca should not interfere with the generation of high l y radiopure 47 Sc in future work, the radio nuclidic purity re p orted for 47 Ca does not consider any 45 Ca present in the purified product. 4 .2.6 Stable Elemental Analysis Water Samples 1 - 4 and the final purified 47 Ca solution containing fractions from each s e paration were analyzed using a semi - quantitative ICP - OES met h od to identify and semi - quantify ions above the LOD of the instrument. [68] Samples of 20 0 mL from water sample 1 - 4 were evaporated on a rotary evapor a tor and reconstituted in 10 mL of 1.4% HNO 3 each. The 96 round bottom flasks used for the evaporation were first rinsed with 1.4% HNO 3 and then with MilliQ water twice. Additionally, all the purifi e d 47 Ca samples from the three separation methods were combin e d and a sample was diluted by half for analysis, resulting in a solution of about 1.5 M HCl. The semi - quantitative method used preset calibration information for 69 elements with a one - point ca l ibration at 5 ppm from the following calibration check standa rds: Rare Earths, Precious Metals, Tellurium, Alkaline Earth Non - Transition Elements, and Fluoride Soluble Group. Among these elements were those that would indicate corrosion of the target or m e tal components in the system ( e.g. , Ti, V, Al, Fe, Ni, Cr) an d elements that have been previously identified as common contaminants in the isotope harvesting system [37] ( e.g. , Na, Mg, Ca, Si, Zn). The samples were run with a blank check solution of 1.4% HNO 3 for the concentrated water samples and 1.5 M HCl for the combined, purified 47 Ca fractions. These blank sampl e s served to help set the baseline for stable ions from the ac id content of the samples and for the blank readings from the semi - quantitative method. Elements with readings from the sample above d to be present and were used as a semi - quantitative estimate of the amount present. 4 .3 Re sults and Discussion 4 .3.1 Production of 47 Ca The total activity of 47 Ca measured (decay corrected to the end of the irradiation) was 3.7 ± 0.7 MBq (100 ± 20 Ci). T he errors considered in this measurement include the counting statistics, errors in reporte d gamma - ray branching ratios, and an uncertainty in the total water volume of the system. By far, the dominant factor is the error in the reported branching r atios 97 Table 4 . 1 : 47 Ca Ac tivity Measured in Each Water Sample and Cation Exchange Resin Bed Sample Activity (kBq) Sample Activity (kBq) Wate r sample 1 7(1) Cation Resin 1 3.0(6) X 1 0 2 Water sample 2 16(3) C ation Resin 2 9(2) X 10 2 Water sample 3 25(5) Cation Resin 3 1.2(2) X 1 0 3 Water sample 4 34(7) Cation Resin 4 2.3(5) X 10 2 Water sample 5 42(8) Cation Resin 5 3.1(6) X 10 2 Remaining water 7(1) X 10 2 Anion Resin 5(2) o f 19 - 20% error for the three main gamma rays. Table 4 .1 gives the decay corrected activity for 47 Ca in each of the samples measured. A production rate of 0.020 ± 0.004(4) 47 Ca nuclei produced per incoming 48 Ca nuclei was measured. The reported erro r for this rate is solely from the uncer tainty in the quantification of the total activity of 47 Ca produced. The production rate can also be thought of as a 2.0 ± 0.4 % conversion rate of beam particles to the desir ed nucleus, which is relatively high for a charged particle irradiation. In compa rison, this rate is 10 to 20 times higher than that for 18 F through the routine production route of 18 O(p,n) 18 F. [69] The pre d icted production rates using both PHITS and LISE++ are given in Table 4 .2. These predictions are s ignificantly lower than the production rate measured in this work. This difference demonstrates the impo rtance of measuring the actual production rate of rad i onuclides in the isotope harvesting system as predictions have been found to be inaccurate as seen previously with a 40 Ca beam experiment with a water target at the NSCL. [70] 98 Table 4 . 2 : Pr edicted and Measured Production Rates of 47 Ca in Isotope Ha rvesting Water Target with a 140 MeV/nucleon 48 Ca Beam The production rate is given as the percent of beam particles co n verted to 47 Ca. The measured production rate should be the same for higher intensity irradiations a nticipated at the NSCL and FRIB. This allows for estimates for the activity anticipated at th ese higher i ntensities and for more detailed safety and experiment al planning f o r future isotope harvesting of 47 Ca. Approximately 4.8 GBq (130 mCi) would be expect ed at the end of a 120 - hour ( i.e. , 5 - day) irradiation with a 140 MeV/nucleon 80 pnA 48 Ca beam, assuming 90% of the primary beam is directed to the isotope har vesting beam b locker. Since 47 Ca would be produced as a byproduct of the NSCL experimental program , this estimate uses the average length of a typical nuclear physics experiment at the NSCL and the stan dard settings for the 48 Ca beam available at this facility . Withou t any dedicated beamtime or additional use of enriched 48 Ca, a significant supply of 47 Ca could be p roduced for research purposes during normal NSCL operations. The measured production rate can be extended to the 48 Ca beam at FRIB as an underestimation of t h e potential production of 47 Ca. At FRIB, the 48 Ca beam that reaches the isotope harvesting beam du mp will have an estimated energy of 189 MeV/nucle on. With a Source of Production Rate Beam Energy at Isotope Harvesting Target (MeV/nucleon) Production Rate (%) Experimentally me asured 140 2.0 ± 0.4 Predicted with PHITS 140 1.19 Predicted with LISE++ 140 1.03 Predicted with LISE++ 189 1.72 99 higher energy beam, a larger fraction of the beam particles undergo fragmentation reactions, res u l ting in a higher production rate of 47 Ca. For example, the predicted rates from LISE++ for the two different beam energies are given in Table 3.2. [29] While it has been noted that the absolute production ra tes predicted by LISE++ differs from the exp erimentally measured rate s , this program provide s reliable relative trends. Using the production rate measured i n this experiment, a 1 - day irradiation of the isotope harvesting beam dump at FRIB full beam power ( 189 MeV/nucleon 30 p A 48 Ca beam, 86% primar y beam transmission to beam dump) would produce >520 GBq (14 Ci) of 47 Ca. As with the 47 Ca produced at the NSCL, the 47 Ca production at FRIB will occur simultaneously with the nuclear physics program as the unuse d primary beam from these experiments is sto pped in an aqueous beam dump. [39] 4 .3.2 Collection and Sample Processing The efficiency with which 47 Ca was collected from the system o n Resins 1 - 3 with ap proximately 5.5 hours of water flow on the day of the irradiation was found to be 65 ± 1%. Two additional resin beds were used subsequently to increase the collected fraction, resulting in 82 ± 3% of the 47 Ca collected on five resi n bed s . The uncertainties considered for these ef ficiencies result from counting statistics and an uncertainty in the volume of water in the system. Offline experiments have demonstrated that increasing the flow through the ion exchange resins can increase the overall collection e fficiency . While there i s a lower collection the column. [36] For futur e experiments in which the half - life of the radionuclide of interest is short, the flow rate through ion exchange resins can be increased to expedite the collection process. 100 In addition to 47 Ca, other cationic radionuclides were collected on the cation exc hange resin beds. The activities on these resins decay corrected to when they were removed from the system are given in Table 4 .3. Short lived radionuclides were not observed on Resin 3 because it was removed hours after the irradiation ended. Therefore, t he activities on Resins 1 and 2 give a view of the radionuclides and their activities that would be encountered soon after an irradiation and the activities on Resin 3 represent the activities that would be present after a 21 - hour cool down period. Using 3 M HNO 3 , an average of 96 ± 1% (n=4) of the 47 Ca collected on the cation exchange resins was removed with 50 - 70 mL. The highest removal rate was observed when 70 mL were used to remove 99.8 ± 0.7% of the 47 Ca from Resin 1 and the lowest was observed when 55 mL were used to remove 89 ± 2% from Resin 4. The sodium, magnesium, and potassium isotopes were entirely eluted from the collection resins, while the scandium isotopes were eluted to a lesser extent (Table 4 .3). The first 50 - 55 mL of 3 M HNO 3 used to re move 47 Ca from each column were used in the separations. When more than this volume was used in the elution from the collection resins, the last few milliliters contained a low activity of 47 Ca due to tailing elution behavior. For the fractions evaporated to dryness for separation methods 2 and 3, the activity was reconstituted in the proper matrix for each load solution with a high yield: >99% for separation method 2 and >98% for separation method 3. Since no 47 Ca was detected on the flasks after the trans fer, these yields are lower limits found using the LOD in the gamma spectra at the 47 Ca characteristic energies. 101 Table 4 . 3 : Quantification of Radionuclides Collected on Cation Exchange Resins 1 - 3 4 .3.3 Purification of 47 Ca 4 .3.3.1 Separation Methods Each separation method was pe rformed three times to confirm the elution profiles of all radionuclides involved and the separation yield and radionuclidic purity for 47 Ca. A representative elution profile for the replicate with the most finely collected fractions for each method is g iv en in Figure 4 .5 with d etails of the fraction volumes and compositions given in Tables 4 .4 to 4 .6 for Separation Methods 1 to 3, respectively. The error bars in Figure 4 .5 result only from statistical uncertainties in the detection method. 102 Depending on th e time each separation was performed, a slightly different mix of radionuclides was identified due to their half - lives and production rates. For example, separation methods 2 and 3 required optimization, so 42 K was not observed in the final elution profi le s for either of these methods. As would be expected, 42 K followed the elution pattern of 43 K in all separations performed in this work, including the separations performed to optimize these two methods. Therefore, it can be conf idently assumed that the e lu tion profile for 42 K is the same as that of 43 K in the final protocols for methods 2 and 3. Conversely, 46 Sc was not observed for any of the three replicates performed for separation method 1 due to the level of activity of othe r radionuclides at the tim e of separation and the relatively long half - life of 46 Sc. However, this scandium isotope should behave identically to 44m Sc, 47 Sc, and 48 Sc and each of these isotopes were found exclusively on the DGA resin after the elution of 4 7 Ca. Two radionuclides ap pe ar during purification as daughters of isotopes produced during the irradiation: 28 Al as the daughter of 28 Mg, and 44 Sc as the daughter of 44m Sc. The half - life of 28 Al (t 1/2 = 2.245 min) is so short that it has an apparent eluti on with 28 Mg in all separa ti ons. Like all the scandium isotopes observed, 44 Sc remained on the column through each of these separations. Therefore, neither of these daughters affected the radionuclidic purity of the final 47 Ca sample. 103 Figure 4 . 5 : Elution Profiles for Separation Methods 1 , 2, and 3 The separation profiles for methods 1 to 3 (1: DGA with HNO 3 /HCl, 2: AG MP - 50 resin with an HCl gradient, and 3: AG MP - 50 resin with HCl/methanol gradient) are sho wn in a to c, respectively . The liquid phase as well as its composition and volume are indicated along the x - axis for each separation through the separation. The error bars are f ro m counting statistics for one replicate of each method and not from deviations across replicates, with large errors resulting from low activities in the samples. 104 Table 4 . 4 : Example Replicate with Separation Me thod 1: DGA with 3 M HNO 3 /3M HCl The recovery percentage i s given for each elution fraction. Two rows at the bottom of the table give the percent that w as hat is blank indicates a value of zero. Liquid Phase Fraction Number Fraction Volume (mL) Volume Sum (mL) Recovery Percent (%) 24 Na 28 Mg 42 K 43 K 47 Ca 44m Sc 47 Sc 48 Sc 3 M HNO 3 1 15 15 100(5) 10 0(12) 100(9) 100(2) 2 7 22 3 M HCl 3 10 32 3.4(5) 4 10 42 97(3) Column 100(17) 100(3) 100(7) Total 100(5) 100(12) 100(9) 100(2) 101(3) 105 Table 4 . 5 : Example Re plicate with Separation Method 2: AG MP - 50 with HCl Gradient The recovery percentage is given for each elution fraction. Two rows at the bottom of the table give the an d the total percent that was eluted from the zero. Liquid Phase Fraction Number Fraction Volume (mL) Volume Sum (mL) Recovery Percent (%) 24 Na 28 Mg 43 K 47 Ca 44m Sc 46 Sc 47 Sc 4 8 S c 0.1 M HCl 1 20 20 2 10 30 2 M HCl 3 5 3 5 100(19) 100(32) 4 4.5 39.5 5 4.5 44 34(2) 6 2 46 40(2) 7 2 48 19(1) 8 3 51 6.3(7) 9 2 53 5 M HCl 10 2.1 55.1 15.8(7) 11 5 01.1 82(2) 12 2.4 62.5 1.7(3) Column 100(5) 100(13) 100(1) 100(3) Total 100(19) 100(32) 100(3) 100(2) 106 Table 4 . 6 : Example Replicate with Separation Method 3: AG MP - 50 with H Cl / Methanol Gradient The recovery percentage is given for each elution fraction. Two rows at the bottom of the table give the percent that was eluted from t h y cell in the table that is blank indicates a value of zero. Liquid Phase Fraction Number Fraction Volume (mL) Volume Sum (mL) Recovery Percent (%) 24 Na 28 Mg 43 K 47 Ca 44m Sc 46 Sc 47 Sc 48 Sc 0.5 M HCl/90% MeOH 1 13 13 2 6 19 3 10 29 4 8 37 2 M HCl/60% MeOH 5 14 51 6(3) 53(19) 6 16 67 57(10) 47(15) 7 13 80 37(9) 8 6 86 2 M HCl/30% MeOH 9 7.5 93.5 7.5(5) 10 4.5 98 19(1) 11 3 101 23(1) 1 2 3 104 25(1) 13 4 108 17(1) 14 4.5 112.5 10(1) 15 4 116.5 1.6(5) 16 3.5 120 4 M HCl 17 13 133 99(4) 18 7 140 1.6(4) Column 100(27) 100(28) 100(2) 100(13) Total 100(14) 100(24) 103(3) 10 0 (4) 107 4 .3.3.2 Separation Yield and Radionuclidic Purity The average separation yields and radionuclidic purities for the three replicates for 47 Ca processed with these separation methods were quite high as shown in Table 4 .7. The only separation tha t d emonstrated less than 100% separation yield was one of the replicates for separation method 3 and even then, it had a 99 ± 2% average separation yie ld with high radio nuclidic purity. The LOD with HPGe measurements for 47 Ca and 43 K for the separation yie ld and radio nuclidic purity, respectively, were found to be within the statistical errors of the final values. Therefore, within the limits of detectio n, these methods resulted in 99 - 100% separation yield with 100% radio nuclidic purity of the 47 Ca recovere d. This high purity indicates that any of the three methods would facilitate the generation of 47 Sc with high radionuclidic purity for radiolabeling ap plications. Figure 4 .6 demonstrates through gamma - ray spectra the radionuclidic purity of the final 47 Ca sa m ple from a replicate separation for Separation Method 1 compared to the radionuclidic mixture in an irradiated water sample from this experiment. Table 4 . 7 : Separation Yield and Radionuclidic Purity of 47 Ca f o r Three Separation Methods Separati on method 1 used DGA resin with HNO 3 /HCl, 2 used AG MP - 50 resin with an HCl gradient, and 3 used AG MP - 50 resin with HCl/methanol gradient. 108 Figure 4 . 6 : 47 Ca Purification Gamma Spectra G amma spectr a for an irradiated water sample (a) and a pu rified sample of 47 Ca after a replicate separation with Separation Method 1 (b) that demonstrate the effectiveness of the purification process including collection with a cation exchange resin, a wait ing period to allow for decay of short - lived cationic ra d ionuclides, and the column based chromatographic separation. 4 .3.4 Comparison of Separation Methods 109 110 4 .3.5 Stable Elemental Analysis T he elements detected with the semi - quantitative ICP - OES method at a concentration of 0.05 ppm or higher in any of the samples are shown in Table 4 .8. The concentrations listed for the combined, purified 47 Ca sample are corrected fo r the dilution made before analysis and are dissolved in approximately 150 mL of 3 - 5 M HCl. The elements detected in the concentrated water sample are only reported for water sample 3 since this sample was the last withdrawn before cation exchange resins w ere put in the system. The values for the water sample are given at the concentrated level (column 6 in Table 4 are also scaled up to account for the total water volume in the system (35 ± 4 L) at the time that water s ample 3 was withdrawn (column 7 in Table 4 Estimated Mass in Total System Water Based on Sample 3 111 Table 4 . 8 : Stable Element Semi - Quantification All values reported in this table are given to only one si gnificant figure with unknown accuracy due to the semi - quantitative nature of the measurement. The measured concentrations for elements in the blank for each sample type was approximately zero ( - 0.03 to 0.03 ppm) except for those listed in the table and ce sium, which read about - 14 ppm for blanks and samples, indicating that the preset calibration was misaligned. A dash signifies that an elemental concentration was < 0.05 ppm for that t he 35 L remaining in the system after water sample 3 was removed. Only small amounts of stable ions were found in the water samples, indicating that the system did not contribute significant amounts of stable ions under these irradiation conditions. This was a concern as a beam of energetic particles creates corrosive radiolysis products such as H + , OH - , HO . , H . , HO 2 . , and H 2 O 2 as it deposits energy and stops in water . [36] While mos t of these radiolysis products recombine rapidly, H 2 O 2 is long - lived and can cau se oxidative damage to metal components exposed to the system water. In this water system, only a few stainless - steel components are in contact with the water and could become a source of Fe, Cr, and Ni. This stable element analysis demonstrates that under the irradiation conditions in this Element Wavelength (nm) Concentration (ppm) Estimated Mass in Total System Water Based on Sample 3 ( g) Blank - 1.5 M HCl Combined Purified 47 Ca samples Blank - 1.4% HNO 3 Concentrated Water Sample #3 Ca 396.847 - 2 - 0.1 200 Ca 422.673 - 3 - 0.2 400 Mg 279.553 - 0.2 - 0.08 200 Na 589.592 0.1 0.2 0.09 0.4 700 Si 251.611 0.7 0.2 - 0.09 400 B 249.77 2 0.06 - - 0.3 200 Fe 238.204 - 0.1 - 0.05 80 Zn 213.857 - 0.3 - 0.3 100 Cu 327.395 - - - 0.05 80 112 experiment, the mass of stable ions that accumulated in the water was low, and therefore, the corrosive effects of radiolysis in the water s ystem were minimal. F igure 4 . 7 : Stable Ions in System Water Compared to Purified 47 Ca Fractions 113 4 .4 Conclusion 114 Chapter 5 : Higher Intensity 1 H Beam Target Dur ability Test 9 A high intensity proton irradiation was performed with the flowing - water isotope harvesting target at the University of Wisconsin - Madison Cyclotron Laboratory to measure the rate of degradation of the Ti64 target shell under high - intensity irradiation conditions. Radionuclides formed in the target shell were measured in the system water as a radiotracer for target degradation. Using a simple, beam intensity dependent model, a corr osi on rate of 1.5E - 6 m/( A*s) was found to match the measure d radiotracer activities at various points in the irradiation. This rate was used to extrapolate the lifetime of future isotope harvesting targets at the NSCL and FRIB, using the areal power density of different ion beams to scale the corrosion rate. 5 .1 Introduction To safely harvest isotopes at the beam intensities available at the NSCL and in the future at FRIB, the second - generation flowing - water isotope harvesting target (see Chapter 2 Section 2 .2.2) was tested under the intense irradiation conditions anticipated . In particular, the durability of the Ti64 alloy (6% aluminum, 4% vanadium, mass balanced by titanium) target shell must be assessed before the harvesting system is used for routine operation . Du ring 9 This chapter draws heavily on the published paper E. Paige Abel, Katharina Domnanich, Colton Kalman, Wes Walker, Jonathan W . Engle, Todd E. Barnhart, Greg Severin; Durability test of a flowing - water target for isotope harvesting, Nuclear Inst. And Methods in Physics Research B , 2020 , 478 , 34 - 45. 115 ir radiation, three main conditions might lead to degradation of the Ti64 target face : 1) displacements of the atoms caused by the energetic beam as it passes through the Ti64 layer , 2) corrosion by radiolysis products in the target water produced as t he acce lerated beam deposits energy in the water, and 3) erosion by the rapid flow of water over the interior surface of the Ti64 layer . The first two of these conditions may become more problematic as the beam intensity increases. Previous isotope harves ting ex periments, summarized in Table 5 .1, have been performed at the NSCL using a target shell in part or entirely made of Ti64 using 40 Ca 20+ , 48 Ca 20+ , 78 Kr 35+ , and 78 Kr 36+ beams with power depositions from 0.42 to 22.2 W (see Chapter 3 and 4 for a discus sion of the 40 Ca 20+ and 48 Ca 20+ beam experiments, respectively) [36,37] . An increasingly intense fast, heavy ion beam was used in these experiments with no observed degradation, b ut even still, the target will be expected to receive much higher beam intensities when installed at the beam blocker position at the NSCL and in the future, at FRIB. For comparison, the present proton irradiation which is intended to replicate the higher power depositions expected at the NSCL and FRIB is summarized in the bottom row of Table 5 .1. There are published techniques for investigating irr adiation accelerated corrosion at a metal - water interface [74,75] . For example, Wang and Was studied the interfa ce between a zirconium alloy and super - heated water by irradiation with protons by using focused ion beam mill ing to create a cross - sectional view of the sample and techniques in transmission electron microscopy to characterize the sample [75] . After irradiation, surface alterations and corrosion rates were determined electrochemically. For the case of the NSCL beam - blocker, and eventually the beam dump at FRIB, the design of the target includes a fully enclosed interior 116 surface, making the use of similar post - irradiation analysis impossible without destruction of the target. To approach the high - power density scenario expected at NSCL and FRIB, and to assess material degradation non - invasively, t he harvesting beam blocker was temporarily installed at the University of Wisconsin - Madison Cyclotron Laboratory. There, a GE PETtrace cyclotron was used to implant a 16 MeV proton beam at currents up to 34 µA into the Ti64 flowing - water target and isotop e harvesting system. The purpose of the irradiation was to measure the degradation rate of the target material in a comparable power density setting to that expected at the heavy ion facilities by observing corrosion products in the water . The UW proton be a m deposited 540 W of total beam power into the target shell and the interior water at the maximum beam intensity in this experiment . This power deposition is comparable to that for beams potentially implanted in the beam blocker at the NSCL ( e.g. 540 W of power would be deposited in the beam blocker by a typical 80 pnA, 140 MeV/nucleon 48 Ca beam and 517 W of power would be deposit ed by a typical 45 pnA, 140 MeV/nucleon 82 Se beam). The results of t he present experiment as well as the implications for future isotope harvesting at FRIB are discussed here. Table 5 . 1 : Summary of Experiments Performed with Ti64 Target Material and Isotope Harvesting Water System Experiment Index Particle Beam Energy (MeV/nucleon) A v erage Intensity (pnA) Average Power Deposition (W) Duration (h) 1 40 Ca 20+ 140 0.42 2.4 3.5 2 48 Ca 20+ 140 0.92 6.2 5.1 3 78 Kr 35+ 150 1.6 18.7 5.8 4 78 Kr 36+ 150 1.9 22.2 7.6 5 1 H + 16 2.1 x 10 4 337 3.3 117 5 .2 Experimental Methods 5 .2.1 Experiment Design At the University of Wisconsin - Madison Cyclotron Laboratory, a 16 MeV proton beam was impinged on the flowing - to 34 µA over the course of approximately 3.3 hours of irradiation time. The irradia tion was broken into four irradiation periods (referred to as periods 1, 2 , 3, and 4) as shown in Figure 5 .1 and the beam intensity was increased with each successive period. With the implantation of an energetic proton beam into the target, both reaction s with the interior water and the front layer of the Ti64 target shell were exp ected as depicted in Figure 5 .2 . In the water, 16 13 N and 18 O(p,n) 18 F reactions were expected to be the most Figure 5 . 1 : Timeline for P roto n I rradiation The timeline shows the average beam intensities of each irradiation period and when water sample s, ion exchange resins, and boric acid traps were removed from the system. The total elapsed time from the start of the irradiation to collection of the final columns and water samples was approximately 24 h. 118 significant reactions . The target shell mate rial , which was made mainly of natural Ti and smaller amounts of Al and V, was expected to undergo 48 Ti(p,n) 48 V and 51 V(p,n) 51 Cr reactions which pr oduce a sufficient activity of long - lived products for detection. For the purposes of testing the durability of t he target material, both 48 V (t 1/2 = 16.0 d) and 51 Cr (t 1/2 = 27.7 d) [28,76] produced in the front Ti64 layer served as radiotracers for target d egra dation. Any activity from these radionuclides found in the water w ould indicate the removal of material from the front layer of the target shel l, as this is the only location they were produced. Each of these radionuclides w as observed and the methods used for detecting and quantifying these radionuclides will be described in the Radiation Measurements section below. Beam intensity values were manually re corded throughout the experiment (measured with a Keithly 414A ammeter connected to the tar get ). Th e design of the target system did not allow for suppression of the current measurement, meaning that the recorded values were Figure 5 . 2 : Schematic Cross Section of the Isotope Harvesting Target The layers of the isotope harvesting flowing - water target are shown with the left side of the picture corresponding to the front of the target. The energetics of the proton beam are given for the first Ti64 yer. Additi onally, 48 V and 51 Cr are shown as products in the target face and 13 N and 18 F as products in the first water layer. 119 expected to be proportional to, but not equivalent to, the true beam current. A scaling factor of 0.67(3) between the true current and the r ecorded beam current ( i.e. , was determined by benchmarking the total production of 48 V and 51 Cr against the expected production computed using data from the IAEA monitor reaction cross - section database [69,77] . The measured activi ties of 13 N and 18 F that were produced in the water were found to be consistent with the scaling factor, but due to a large systematic uncertainty that arose from an uncertainty in the total water volume, and t he lack of uncertainties in the IAEA evaluatio n, those values were not used to calculate the scaling factor [69] . By experim ental design, the 3.4 to 34 µA range of beam c urrent was chosen to simulate the beam power deposited in the target shell and water by the highest power beams available at the NSCL , such as the 140 MeV/nucleon 82 Se beam at 45 pnA . Based on SRIM calculations [49] ,thi s beam would deposit approximately 67 W in the first layer of Ti64 in the target shell and 450 W in the interior volume of water. A 16 MeV proton beam between 9 and 55 µA can simulate th is power dep osition . The irradiation started with an average beam curr ent of 5.6 µA for the first irradiation period and the current was increased gradually up to 34 µA for the last irradiation period, achieving a pproximately two - thirds the intended high intensity bea m. The implications of these power deposition comparisons will be discussed later with regards to quantify ing the degradation of the target. The slow increase in beam current allowed for multiple measurements in the system, such as the activity of radionu clid ic products and the conductivity of the water , to be c orrelated with the beam intensity. To make measurements of the radioactivity in the water , samples of the system water were taken, and the ion exchange columns and the boric acid trap were 120 exchanged for fresh components between each run segment . Figure 5 .1 shows the timeline of when water samples, ion exchange columns, and boric acid traps were removed from the system during the experiment. 5 .2.2 Quantification of Radionuclides 5 .2.2.1 Radiation Measurements for the Quantification of Radionuclides The acti vit ies of 13 N and 18 F in the water samples, ion exchange columns, and boric acid traps removed from the system w ere measured with Capintec CRC - 15R ionization chambers calibrated for qua ntification of their 511 keV gamma rays. The sample positions were not changed as activity measurements were recorded over time, allowing for a decay curve analysis for the quantification of 13 N and 18 F. No activity was observed on the boric acid trap thro ughout the experiment. Equation 5 .1 was used for the determination of the activity of 13 N and 18 F on the anion exchange columns and the water samples: ( 5 .1 ) where and were fitted variables for the initial ac tivity of 13 N and 18 1 and 2 are the decay constants of 13 N and 18 F, respectively; b represented the backgrou nd which was restricted to be greater than or equal to zero; and t is the time at which a total activity value, , was recorde d. The data for w ater sample 1 was fitted with the background factor set to zero . The activity found in each water sample wa s scaled by a factor comparing the volume of the water sample to the total system volume ( i.e. , approximately 50 mL samples were 121 measured and the activity was scaled to account for about 36 L in the water system.) to produce an approximate activity in the total water volume. Only 13 N was found on the cation column. This was verified by fitting the ionization chamber values for the cation columns with Equation 5 .2 : ( 5 .2 ) where was a fitted variable for the initial activity of 13 N, is the decay constant of 13 N, b represented the background which was restricted to be greater than or equal to zero, and t is the time at which a tot al activity value, , was recorded. The activity found on the ion exchange columns and in the total volume of water were used to find a total activity of 13 N or 18 F in the system at the end of each irradiation period. Longer - lived radionuclides such as 48 V and 51 Cr were quantified by gamma spectrometry using two energy, efficiency, and peak shape calibrated Al - windowed Ortec HPGe detectors (measured resolution 1.8 keV FWHM at 1333 keV). The activities measured in each water sample were scaled to acco unt for the total system water volume as described previously. Spectra were taken starting the day after the irradiation to reduce the background from the shorter - lived co - produced radionuclides. In particular, 48 V was quantified with its characteristic ga mma rays at 983.525 keV (99.98(4)%) and 1312.106 keV (98.2(3)%) [28] and 51 Cr was quantified with its 320.08 keV (9.910(10)%) gamma ray [76 ] . Two HPGe detectors were used for quantification at the Cyclotron Lab. Since detector 1 was more well characterized than detector 2, spectra from detector 1 were preferentially used in the quantification. When 122 spectra from detector 2 were the only data available for a sample, an additi onal 10% error was added in quadrature with the counting error. To improve the quantification, spectra were taken of the water samples over long time periods within a month of the irradiation. These spectra were taken in t he Department of Chemistry at Mic higan State University, using a Canberra BEGe Gamma - ray Detector (BE2020). The anion exchange columns were also measured for long counting times about a month and a half after the irradiation at the Cyclotron Lab in Madison to better quantify 51 Cr. The tar get shell was analyzed using gamma spectrometry three and a half months after the irradiation due to the high level of long - lived activity produced in the front Ti64 layer of the target. Analysis of the spectra taken at th e Cyclotron Lab were performed by hand and involved finding the peak sum and subtracting an average baseline from the sum. The efficiencies used were calculated from existing characterization s of the detectors. The spectra taken at Michigan State Universit y were analyzed using Genie 2000 software (Mirion Technologies), which performed peak summing, baseline corrections, and efficiency calculations also based on existing calibrations. Table 5 . 2 : Estimated Produ ction Rates for Radionuclides Produced in Water and Target Shell Radionucl ide Estimated Production Rate (product nuclei/incoming beam particle) 13 N 4.05E - 5 18 F 1.66E - 6 48 V 9.86E - 4 51 Cr 5.64E - 5 123 5 .2.2.2 Estimated Production of Radionuclides Estimated production rates of 13 N, 18 F, 48 V, and 51 Cr were c alculated using the recommended cross sections from the IAEA evaluations for the following reactions, respectively: 16 13 N, 18 O(p,n) 18 F, nat Ti(p,x) 48 V, and 51 V(p,n) 51 Cr. These cross sections were used in Equation 5 .3 : ( 5 .3 ) where (E) is the cross section of the reaction at energy, E; S p (E) is the stop ping power of the projectile at energy, E; is the density of the target nucle i ; M is the molar mass of the target nucleus; and N A energy was estimated using SRIM and range d from 16 to 9.7 MeV in the target shell materia l and from 9.7 to 0 MeV in the water layer. The production rates in Table 5 .2 were found using these parameters and were used with t he recorded beam current in Equation 5 .4 to find estimated activities produced, A: ( 5 .4 ) whe re lambda is the decay constant of the radionuclide and t is the time over which the beam current was I(t). 5 .2.3 Estimating Activity of 48 V and 51 Cr in the System Water from Nuclear Recoil A study on recoil range d istributions by Alexan der and Sisson demonstrated that the total momentum of a projectile is transferred to the compound nucleus formed in a nuclear r eaction [78] . Due to this full momentum transfer and subsequent symmetric particle emission, Equation 5 .5 describes the avera ge recoil energy of the resulting product: 124 ( 5 . 5 ) where E and A repre sent the energy and mass, respectively, and the subscripts R, b, and T represent the recoil, the projectile, and the target nuclei , respectively. This process and the assumptions involved are depicted schematically in Figure 5 . 3 . Usin g stopping power value s from SRIM to estimate E b in Ti64, the recoil energ ies of 4 9 V and 5 2 Cr ( i.e. , the compound nuclei for the reactions 48 Ti(p,n) 48 V and 51 V(p,n) 51 Cr, respectively) w ere estimated with Equation 5 .5 in the target shell at 1 m intervals. The recoil energ ies are predicted to be approximately 199 keV for 4 9 V and 187 keV for 5 2 Cr in the last 1 m of the target shell where recoil escape was able to occur ( i.e. , the remaining layer of Ti64 was small enough that these recoil energ ies moved some of the compou nd nuclei into the water layer) . T he percent of 4 9 V and 5 2 Cr recoils that would travel through the remaining target shell material and into the water layer was evaluated in steps of every 0.001 m in the last 0.2 m of Figur e 5 . 3 : Depiction of Recoil and Production of Radionuclides. The reaction between a projectile and a target nucleus forms a compound nucleus moving in the same direction as the projectile with an energy descr ibed by Equation 5 .5. The compound nucleus emits a particle in any directi on, forming the product of the nucl ear reaction. This symmetric distribution statistically results in no net shift in recoil range distribution. 125 the shell based on the following in f ormation: the average range and range distribution of the recoil ing compound nucleus in Ti64 at the estimated recoil energies and a standard normal distribution table. A Gaussian distribution was fitted to the range distribution of these recoil ing compoun d nuclei . The resulting mean ( ) and standard deviation ( ) were used to standardize the distance ( ) a recoiled particle in each 0.001 m layer would travel to exit the target shell: ( 5 .6) Then , the probability of a recoiled particle traveling through the r emaining standardized distance ( ) was found using a standard is depicted in Figure 5 . 4 . The final product nucleus is produced when the compound nucleus sym metrically ( i.e. , isotropically) emits a particle as shown in the right half of Figure 5 .3. Due to Figure 5 . 4 : Depiction of Recoil Fraction Estimation Projectiles are shown traveling to the right through t he last 1 um of Ti64 an d into the wate r layer. When the projectiles react with the target nuclei , compound nuclei are formed and continue to travel to the right. The red gau ssian curves depict the recoil range of the compoun d nuclei. The fraction of red curve overlapping with the water layer i s proportional to the fraction of product nuclei that recoil into the water layer. 126 this symmetric emission, the position of the recoiled compound nucleus estimated as described previously repr esents the average position of the product nucleus. IAEA eval uated cross section valu es for nat Ti(p,x) 48 V and 51 V(p,n) 51 Cr were used to find production rates at 0.001 um intervals in the last 0.2 um of the target shell. The probability of recoil ing compo und nucleus Ti64 tar get layer that was found fo r each of these interval s the recoil ing compound nuclei was found for each reaction by adding the rates at each laye r. Using the beam current values described in the Materials an d Methods: Experimental Design S ection 5 .2.1 , the activity of 48 V and 51 Cr in the system from recoil ing particles traveling out of the target s hell was estimated and subsequently subtracted from the total activity measured in the water for each irradiation period. 5 .2.4 Degradation of the Target A model was developed to describe the degradation of the first Ti64 layer of the target based on the measurements of the 48 V and 51 Cr produced in this l ayer and found in the water system. The amount of 48 V or 51 Cr which had escaped from the first Ti64 targ et layer into the water at each time point was the sum of activity on the ion exchange resins and the scaled activity in the water samples. These activi ties were corrected for the estimated activity in the water re sulting from nuclear recoil following the 48 Ti(p,n) 48 V and 51 V(p,n) 51 Cr reactions in the target shell. Therefore, the final activities used in this model represent the activity that resulted fro m degradation of the target material. This calculation was car ried out with 1 - to 2 - minute time steps an d the scaled beam intensities described in the Experimental Design section. 127 The measured activity of 48 V and 51 Cr in the target and the beam current du ring the irradiation were used to calculate a production rate for these radionuclides in th e first Ti64 target layer . This production rate was used to find activities for 48 V and 51 Cr in the Ti64 target layer at each time step in these calculations using E quation 5 .7 : ( 5 .7 ) where P T (t i ) is the production of 48 V in the Ti64 target layer over the time step starting at time t i , I(t i ) is the beam intensity at t i (pps), r is the constant production rate o f 48 V or 51 Cr (particles produced per each incom ing proton), is the decay constant of 48 V or 51 Cr, and t i+1 is the beginning of the next time step. The total activity in the first Ti64 target layer at each time step, A T (t i ), was found by adding the act ivity produced at t i (Equation 5 .7 ) to a decay c orrected value for the accumulated activit y in th is target layer at t i - 1 , the previous time step, as shown in Equation 5 .8 : ( 5 .8 ). The value produced with Equation 5 . 8 for the last time step in the irradiation is equal to the amount of 48 V or 51 Cr produced by the end of the irradiation. The amount of 48 V or 51 Cr that degraded from the first Ti64 target layer at each st ep, t i , was assumed to be a function of the acti vit y of each radionuclide in the target face at that time step. Equation 5 .9 describes the way the target is thought to degrade over time: ( 5 .9 ) 128 where D T (t i ) is the activity of 48 V or 51 Cr that was removed from the Ti64 target layer over the time step starting at time t i , I(t i ) is the beam intensity at time t i , and T is t he thickness of the first Ti64 layer . This model uses one variable, C, which is the corrosion rate of the first Ti64 layer in units of degraded this model are presented in Section 5 .3 .2.1 . The area of the proton beam strike and the power deposited per integrated beam current are two factors that were used to scale the corrosion rate resulting from Equation 5 .9 to the dimensions of volume per energy deposited (cm 3 /J) . These units allow the corrosion rate to be rescaled to other irradiation conditions ( i.e. , different beam strike areas and power depositions with a range of beams types and intensities) expected at the NSCL and FRIB . The scaled corrosion rate from this experiment as well a s extrapolations for other beam conditions are discussed in Section 5 .3.2 .2 . The accumulated activity of 48 V or 51 Cr in the system, A s , was found in a similar way to the total activity in the front Ti64 target layer . Equation 5 .10 shows how the accumulatio n occurred at each time step: ( 5 .10 ). The Microsoft Excel solver plug - in was used to a find value for C in Equation 5 . 9 that minimizes the X 2 differ ence between the measured activity of 48 V and 51 Cr in the system at the end of each irradiation period and the value produced by Equation 5 .10 at these times. The results of this model are presented in Section 5 .3.2.1 . 129 5 . 3 Results and Discussion 5 . 3.1 Quantification of Radionuclides Produced in Water Target The total activity of 13 N and 18 F measured in the water system and predicted activities based on the scaled beam current after each irradiation period are given in Table 5 .3. Th e predicted activities are lower than the measured activity in each case. Howev er, the predicted activities for 13 N are within the error of the measured activity. The larger discrepancy in the 18 F data most likely results from two factors: the much lower 18 F activity compared to the 13 N activity in each sample and the large number of resonances in the cross section for the 18 O(p,n) 18 F . The low 18 F activity was a small component in the fit used for the decay curve analysis, so the uncertainty in the fitted activity was likely underestimated. Additionally, the IAEA rec ommended cross se ctions were used to find a predicted activity [69] . These recommended values do not sufficiently represent a ll observed reson ances, making it possible that the predicted activity is an under estimation. Finally, uncertainty information for the IAEA Table 5 . 3 : 13 N and 18 F Activities Produced in the Flowing - Water Target The values giv en in the last column are ratios of the measured to predicted activities after each irradiation period. Radionuclide Irradiation Period Tota l Activi ty Produced after Each Irradiation Period(mCi) Measured Predicted Ratio 13 N 1 60(10) 50 1.29 2 150(30) 130 1.14 3 200(30) 180 1.09 4 250(40) 220 1.11 18 F 1 1.0(2) 0.5 1.91 2 2.0(3) 1.5 1.33 3 2.8(3) 2.0 1.42 4 5.8(8) 4.2 1.36 130 recommended values were not available for either 13 N or 18 F, so it is unclear how precise the predicted values are . Regardless of these relatively small discrepancies, the level of these radionucli des produced in the water provide information about the conditions in the interior water layer. Due to the high LET of the protons , the temperature of the Ti64 front target l a yer has been predicted to be much higher than the boiling temperature of the wate r in the system. This condition introduces the possibility of partial or total vaporization of the 2 mm layer of water between the front Ti64 shell and the internal Ti64 ba ffl e depicted in Figure 5 . 5 . A significant decrease in the expected producti on of 13 N and 18 F could indicate that the water had vaporized to an extent that the beam passed through the water layer and was implanted in the second Figure 5 . 5 : Proton Beam Implanted in Isotope Harvesting Target The layers of the isotope harvesting target are shown with the front Ti64 target layer on the left followed by a water layer (darker blue), an interior layer of Ti64, and a second large water layer . The 16 MeV protons are expected to deposit approxi mately 6.3 MeV of energy in the front Ti64 layer and the remainder in the first water layer with an estimated range of 1.1 mm. However, if the average density of the water layer were to decrease due to lo cal boiling (shown in the lighter blue), the range o f the p roton through the water could exceed 2 mm and some of the energy of the protons could be implanted in the second Ti64 layer. 131 l ayer of Ti64. However, a compa rison betwee n the measured and predicted activities for these radionuclid es produced in the water interior of the target does not show a decrease in production , indicating that the proton beam was most likely implanted entirely in the water layer. T his means that the co r rosion of the front Ti64 target layer was assisted by real istic levels of radiolysis products from th e proton beam. 5 .3.2 Degradation of the Target 5 .3.2.1 Estimating the Corrosion Rate The model created to describe the degradation of the front Ti64 tar ge t layer use d the fitted constant corrosion rate introduced above . Additionally, the amount of 48 V that was removed from this target layer at each time step was assumed to be proportional to the beam intensity and inversely proportional to the area of the b e am spot. These relationships are Table 5 . 4 : 48 V and 51 Cr Activity in the Water The total measured, recoiled, recoil co rrected, and modeled activities (left to right) are given for the end of each run segme nt. For 48 V, the fitted data fr om Equation 5 . 10 matched the measu red data relatively well with X 2 (3, n=4) = 1.71. Radionuclide Irradiation period Total Activity (µCi) Measured in System Estimated from Recoil into Water Estimated from Target Shell Degradation (Recoil Corrected) P redicted in System by Degradation Model 48 V 1 0.33(6) 0.35 - 0.01(6) 0.04 2 1.7(2) 1.1 0.6(2) 0.4 3 4.0(4) 2.4 1.6(5) 2.0 4 11.7(3) 4.49 7.2(7) 7.0 51 Cr 1 0.028(2) 0.014 0.013(2) 2 0.082(8) 0.046 0.036(5) 3 0.15 7(9) 0.097 0.060(6) 4 0.54(4) 0.18 0.35(4) 0.35 132 intuitive, fit the experimental data well, and allow for the resulting corrosion rate constant to be sc aled by the areal beam power of other particle beams for estimates of future degradation. The measu red activities of 4 8 V and 51 Cr in the water system were corrected for estimated contributions from nuclear recoil s from the front Ti64 target shell layer . A bout 39% of the 48 V and 34% of the 51 Cr measured in the water system were attributed to recoiled pro duct 0.110 µm, respectively. The recoil - corrected values in Table 5 .4 were used in the target shell corrosion model. The optimized results from Equation 5 . 10 for the recoil - co rrec ted 48 V activit ies are shown in Figure 5 . 6 . The modeled values all fall within the error for each measured value and the overall fit has a X 2 (3, n=4) of 1.71. The resulting corrosion rate constant as defined in Equation 5.9 is 1.5E - 6 µm/(µA s). The me as ur ed activities of 51 Cr in the system were much lower than those of 48 V due to a low er target nucleus density in the Ti64 target shell. Additionally, the highest intensity gamma ray for 51 Cr is a 320 keV gamma ray with a branching ratio of only 9.90% [76] . Together these factors resulted in 51 Cr present below the detection limits in the 200 - 300 mL water samples withd ra wn for HPGe spectrometry. Spectra were collected a few weeks af ter the irradiation for live times on the order of days. This measurement was supported by estimating the limit of detection of these spectra a t 320 keV as the uncertainty in the background co un t ing rate in this energy region. Then, the amount of 51 Cr in t he total water volume was estimated by assuming the ratio of activity in the water and on the anion exchange column followed that of 48 V. For all of the water samples, this amount of 51 Cr pr ed ic ted in the water fell below or at the limit of detection. 133 Figure 5 . 6 : Activ ity of 48 V in the System Th e fit results of Equation 5.10 are plotted every 1 to 2 minutes throughout the irradiation in the s olid navy - blue line. These results match the measured data points, plotted in orange, within the estimated errors. Due to this uncertainty in the total 51 Cr in the water system throughout the irradiation, only the total activity on the four anion exchang e columns corresponding to the total activity removed fr om the front Ti64 target layer at the end of the irradiation was used in the degradation model. This assumption that the total 51 Cr activity was represented in the activity measured on the four anion exc hange columns is supported by the very low relative a ctivity of 48 V in the water at the end of the collection period . The system water was passed over the fourth ion exchange resins overnight to remove as much 48 V and 51 Cr from the water as possible. Th i s resulted in 98(3)% of the total 48 V being collected on the four sets of ion exchange resins. Assuming that 51 Cr behaves similarly to 48 V in the system, the sum of activity on these four columns was recoil - corrected and used in the degradation model as t h e total amount 134 degraded fro m the Ti64 front target layer . The model reproduced this value for 51 Cr with a corrosion rate of 2.3E - 6 µm/(µA s). This value is within a factor of two of that found with 48 V, verifying the reproducibility of the estimated corro s io n rate. 5 .3.2.2 Extrapolating Corrosion Rate to Predict Target Lifetimes When extrapolating the corrosion constant measured in this proton irradiation to that for other beams, there are a few important factors to consider. It is particularly important t o take into conside ration the area of the target shell that will be irradiated when making estimates for isotope harvesting at FRIB. Figure 5 . 7 shows the beam power that has been used in previous experiments, a full power 48 Ca beam at the NSCL, and beam p o we rs that are expected at FRIB during the ramp up phase and full power operation. This figure also shows how this comparison between the experi ments changes when we consider the total areal Figure 5 . 7 : To tal Power a nd Areal Power for Various Beams Points labelled 1 to 4 in this figure correspond to the experiments listed in Table 5 .1. Point 5 shows the pro ton irradiation described here and point 6 is given for the experiment s with a 48 Ca beam (140 MeV/nucl eon , 80 pnA ) described in Chapter 6 and 7 . The ramp up of FRIB is represented by the points collectively labeled 7 and point 8 shows FRIB at full beam power. The total power of these beams is shown as pink crosses and the total areal power of these beams i s given in blue triangles. 135 powe r (W/cm 2 ) instead of simply the total power (W). Although FRIB will have orders of magnitude higher beam powers (through both higher energy and intensity beams), this power is spread over a much larger are a on the proposed t arget for isotope harvesting [39] . This means, in terms of total areal power deposited in a target, the present proton irradiation is comparable in magnitude to future full power F RIB irradiations. Additionall y, both the amount of power deposited in the front Ti64 target layer , leading to high temperatures and atom displacements in the target shell, and the amount of power deposited in the water in contact with the inner surface of the Ti64 target layer , produc in g high temperatures and a corrosive radiolysis environment in the water, are instrumental in the degradation of the front target shell layer . Since it is unclear how to scale the degradation rate based on these factors, thre e estimates were made based on t he following beam power scenarios for degradation with future irradiations: 1) using the beam power deposited in the front layer of the targe t shell , 2) using the beam power deposited in the front layer of the target shell a nd the 2 mm water layer, and 3 ) using the beam power deposited in the last 0.1 mm of the front layer of the target shell and in the first 0.1 mm layer of water in contact with the inner Table 5 . 5 : Scaled Co rr osion Rates for Different Beam P ower Considerations The corrosion rate is given for the three different beam power scenarios give n in the text in which different regions of power deposition in the target are considered. Beam Power Scenario Scaled Corros i on Rate (cm 3 /J) Number Descrip ti on of Power Deposition 1 In the first layer of Ti64 1.7E - 11 2 In the first layer of Ti64 and 2 mm water layer 6.5E - 12 3 At the interface between the first layer of Ti64 and the 2 mm water layer 6.6E - 11 136 surface of the f ront target shell layer . These th ree estimates provided different power depositions per integrated beam current. The results are given in Table 5 .5 and 5 .6 for these three scenarios. The time estimates in Table 5 .6 are given as the irradiation time until 5 0 µm or approximately 10% of t he front Ti64 target shell layer is degraded with various beams at the NSCL and FRIB. The 48 Ca and 82 Se beams were selected because they have the highest total power deposition and the highest powe r deposition in the front lay er of the target shell of the bea ms available at the NSCL, respectively. The NSCL beam blocker design used for the estimated target lifetimes ( shown in Figure 5 .2 and 5 .5 and described in detail in Chapter 2 Section 2 .2.2) inv olves a double - walled design an d a 2 mm water gap between the walls. For FRIB, the 40 - kW ramp up beam intensity and the 400 - kW full beam power level are considered for 48 Ca, 82 Se, and 238 U. The first stage of beam intensity a t FRIB will use a first - gener ation beam dump with a single - wal led target: 1 mm layer of Ti64 in contact with a large volume of flowing water. The higher beam power stage will use a second - generation beam dump with a double - walled target design, as in the NSCL isotope harvesting targe t system . This second - generation beam dump will differ from the one used at the NSCL in that it is much larger and is a circular rotating target instead of a rectangul ar stationary targ et to maximize the beam strike area at FRIB. It is clear from these es timates that the proton irradiati on demonstrated accelerated degradation compared to the irradiation conditions expected at the NSCL and FRIB to observe degradation ov er this short expe riment. For perspective on the length of calendar time these prediction s correspond to, it has been esti mated that the NSCL produces beam for about 4000 hours each year. For the 48 Ca and 82 Se beams at the NSCL, about 1500 - 2900 hours are 137 Table 5 .6: Estimated Target Durability for the Highest Power Primary Beams at the NSCL an d Equivalent Beams at FRIB The t arget lifetime estimates for the NSCL were made with the NSCL double - walled beam blocker described in the text. For eac h beam at FRIB, the first row with a lower beam intensity assumes the first - generation single - walled bea m dump design will be used and the se cond row with a higher beam intensity assumes the second - generation double - walled beam dump design will be used. Fo r each of these nine irradiation conditions, target lifetime estimates were made using the thr ee beam po wer scenarios described in the text. Facility Nucleus Energy (MeV/ nucleon) Intensity (pps) Total Power in Window Total Power in Window and 2 mm Wa ter Power at Interface of Window and Water Areal Power Deposition (W/cm 2 ) Time until 10% Degraded (hours) Areal Power Deposition (W/cm 2 ) Time until 10% Degraded (hours) Areal Power Deposition (W/cm 2 ) Time until 10% Degraded (hours) UW - M 1 H 16 2.12E+14 306 275 777 275 76 275 NSCL 48 Ca 140 4.99E+11 122 2445 261 2890 28 2689 82 Se 140 2. 81E+11 196 1521 430 1758 45 1662 FRIB 48 Ca 189 1.87E+13 4 19605 7 31840 1 38560 1.87E+14 20 4215 46 4650 5 3953 82 Se 186 1.16E+13 8 10602 12 17193 1 20777 1.16E+14 36 2332 84 2534 10 2169 238 U 154 4.91E+12 27 3148 49 4337 4 5426 4.91E+13 1 19 707 301 709 33 634 138 e stimated for a 10% deg radation in the thickness of the front Ti64 target layer , which corresponds to approximately 4.5 to 8.7 months of operation. 5 .3.2.3 Validity and Limitations of Extrapolation The present experiment used a proton beam t o form long - lived radiotracers se lect ively in the Ti64 target shell. Fast, heavy ion beams available at the NSCL are predicted to produce the same long - lived radionuclides in the shell and water through fragmentation of the beam, making the m more difficult to interpret than the irradiatio n pr esented here. Additionally, proton beams are widely available at the areal power level needed to replicate those anticipated at the NSCL and FRIB. While this experimental design allowed for the measurem ent of target de gradation and extrapolation to ot her irradiation conditions, there are some important factors that affect the validity of this method. Most notably, the interaction of a proton beam with the target system differs greatly from that of fast, heavy ion beams. The main fundamental difference come s from the linear energy transfer (LET) of these different beams. The LET values of a 16 MeV proton, 140 MeV/nucleon 48 Ca, and 140 MeV/nucleon 82 Se beam as they stop in the target were converted to linea r power depositi on values by using beam intensiti es o f 34 µA for the proton beam and the NSCL beam list intensities for 48 Ca and 82 Se. The resulting values as a function of the distance through the isotope harvesting target are plotted in Figure 5 . 8 . Thoug h all these beam conditions have similar total po wer depositions, the 16 MeV proton has a much higher power deposition in the front Ti64 target shell layer and in the adjacent 2 mm water layer than the 140 MeV/nucleon 48 Ca and 82 Se beams due to the signifi cantly shorter r ange of the proton beam. 139 Figur e 5 . 8 : Linear Power Deposition of Various Beams through the Target The linear power deposition (W/mm) is plotted for a 16 MeV, 34 pµA proton beam; a 140 MeV/nucleon, 45 pnA 8 2 Se beam; and a 140 MeV/nucleon, 80 p nA 4 8 Ca beam through the target. The target materials are shown in order: 590 µm of Ti64, 2 mm flowing water, 59 0 µm Ti64, and a large flowing water volume. One practical result of this different linear power depositi on profile for the proton beam is tha t th e front Ti64 target layer most likely reached a very high temperature of several hundred °C range during thi s irradiation. While the bulk water was conditioned with a chiller and maintained at below 30 degrees C, th e water flowing by the front layer of the target shell was subject to a large increase in temperature as a result of heat transfer from the high temperature target shell and the power deposition of approximately 9.7 MeV per proton in this small volume of w ater. It is possible that partial vap oriz ation of the water occurred leading to a lower capacity to c ool the front layer of the target shell . Th e high temperature resulted from trying to reach high areal power levels with a short - range proton beam and is n ot expected to 140 occur with much longer ran ge fast, heavy ion beams. It is known that high temperatures accelerate degradation rates of materials due to changes in the microstructure of material. Therefore, the corrosion rate measured in this experiment is s uspected to be higher than the rate w hich would be present at lower temperature conditions used for f uture isotope harvesting experiments. This means the target lifetimes predicted from the measured degradation rate in this experiment are most likely lower than the actual lifetimes. Another c lear limitation of this experiment is the large difference in ti me between the experimental irradiation period and the predicted lifetime of the target. Extrapolating to much longer time scales introduces significant un certainties to these predictions. One not eworthy uncertainty is that the short time scale of the irra diation presented here was not able to uncover any long - term changes under irradiation conditions ( e.g. , hardening of the target shell material) that could influence the rate of degradation. T he d egradation is assumed to be constant through hundreds to ten s of thousands of hours of irradiation when the degradation rate may change with developing conditions in the target over time. The results of this experi ment have provided confidence in the inte grity of the isotope harvesting system, including the durabi lity of the target. This has enabled full power 48 Ca irradiations of the isotope harvesting target at the NSCL ( see Chapter 6 and 7 ) . Additionally, the lif etimes predicted in Table 5 .6 support the integrity of the target during the first few years of incre asing the beam intensity at FRIB. During these initial years, the state of the isotope harvesting target to be used at FRIB will be monitored ( e.g. , measur ing the thickness of the target face befo re and after long periods of irradiating). These measurement s will benchmark the real corrosion rate with fast, heavy ion beams at FRIB. 141 5 . 4 Conclusion A proton irradiation of the current isotope harvesting system w as performed to test the durability o f th e target under the large power depositions expected in the i sotope harvesting targets at the NSCL and FRIB. Degradation of the first Ti64 target layer was observed and extrapolated to relevant beams that will be use d at the NSCL and FRIB. These estimat es s how that a 10% reduction in the front Ti64 target layer thic kness occurs on the time scale of several months of operation at the NSCL. Estimates for the ramp - up period at FRIB with the first - generation FRIB beam dum p indicate that thousands to tens of thou sands of hours of irradiation can be performed before a 10% reduction in thickness should occur due a much larger surface area that will be irradiated . Shorter time estimates of hundreds to thousands of hours are pr edicted for th e degradation at the fu ll p ower operations of FRIB with a second - generation beam dump. It is strongly suspected that the corrosion rate measured here is higher than that at the NSCL and FRIB, resulting in underestimates of the lifetime of the NSCL isotope harvesting target and a ll g enerations of the FRIB beam dump under the anticipated irrad iation conditions . Measurements will be made of the target degradation as FRIB increases the beam power used in the first few years of operation. Overall, these results indicate that the degra dati on of the target should be manageable obstacles for isotope harvesting at the NSCL and in the future at FRIB. 142 Chapter 6 : 48 Ca Beam Experiment 2: Proof of Concept for 47 Ca/ 47 Sc Generator and 47 Sc Radiolabeling with Isotope Harvested 47 Ca A stepwise i ncreasingly high er intensity 48 Ca irradiation of the flowing - water isotope harvesting target was performed at the NSCL with the target in the beam blocker position in the A1900 fragment separator. The 47 Ca produced in the firs t three lowest intensity irrad iation segm ents was used to fin d a production rate of 0.0 16 7 ( 4 ) 47 Ca nuclei per incident beam particle . The production rates of other cationic radionuclides were also measured to understand the level of activity that would acc umulate on the cation exchange resin beds . Through the remain der of the irradiation, the beam intensity was increased up to a maximum intensity of 80 pnA. A majority of the 47 Ca produced in this irradiation was effectively collected on two cation exchange resin beds in the aqueous chem istry loop of the water system. The first resin bed with the majority of the 47 Ca activity was processed and the eluted activity was used in several steps: purification of 47 Ca, generation of 47 Sc in the purified 47 Ca solution , purification of 47 Sc from th e generator solution , and radio labeling DTPA - TOC with the purified 47 Sc . The generator procedure for the purification of 47 Sc produced a high purity 47 Sc sample in terms of both radionuclidic and stable ion purity, allowing fo r radiolabeling of DTPA - TOC wi th essentia lly 100% radiolabeling yield. Each of these steps demonstrated the proof - of - concept that harvested 47 Ca can be used to generate 47 Sc of sufficient quality for nuclear medicine research . 143 6 .1 Introduction Following the high intensity proton irra diation tha t demonstrat e d the durability of the second - generation isotope harvesting target (see Chapter 2 Section 2 .2.2), th e target and water system w ere used in a higher intensity 48 Ca beam irradiation at the NSCL. The tar g et was installed at the A1900 beam blocke r position (Figure 2 . 3 in Chapter 2 Section 2 .1 ) and was connected to the water system in the final configuration ( Chapter 2 Section 2 .3.3 ) . This experiment was designed to demonstrate both isotope harvesting of 47 C a at the full 48 Ca beam intens ity availab le at the NSCL and to further the purification process investigated in the first 48 Ca irradiation (see Chapter 4 ). The a pproximately 1 . 8 mC i of 47 Ca produced in this irradiation allowed for optimization of the gener ator purification procedure an d the radio labeling of DTPA - TOC ( i.e. , d iethylenetriamine pentaacetate - (Tyr 3 ) - octreotide) with 47 Sc. DTPA - TOC, which is shown in Figure 6.1, consists of an open chain chelator (DTPA) and a biologically active peptide Figure 6 . 1 : Chemical Structure of DTPA - TOC The open chain ligand DTPA is highlighted in blue while the peptide TOC is highlighted in green. 144 that has been shown to interact with somatostatin receptors on the surface of neuroendocrine tumors (TOC) . [79] The methods presented here produced high quality [ 47 Sc] Sc - DTPA - TOC, supporting the feasibility of using isotope ha rvested 47 Sc for preclinical t argeted int ernal radiotherapy research. 6 .2 Materials and Methods 6 .2.1 Materials 6 .2.1.1 Chemicals and Resins Prior to the irradiation, the water in the isotope harvesting system was purified with a mixed bed resin (McMaster - Carr, Filter media PVC water d eionizer) t o lower the conductivity of the water to about 250 nS/cm. Three resins types were used in this work for collection and purification of radionuclides : AG 50Wx8 cation exchange resin (BioRad, mesh size 20 - 50), AG MP - 5 0 cation exchange resin (BioRa d, 100 - 200 mesh size), DGA extraction chromatography resin ( TrisKem International , - tetra - n - octyldigly colamide , normal resin, particle size 50 - 100 ). Both cation exchange resins were pretreated separately in large quantities with the following r inse order: 50 mL of 2 M HCL, 50 mL of 4 M HCL, 50 mL of 6 M HCL, and 100 mL MilliQ water. This procedure was r epeated for a total of 2 complete rinse cycle s, removed ionic impurities, especially metal ions, from the resin , and ensured that the resin was i n the H + fo rm. For all elutions, dilutions, and column separations, metal free hydrochloric acid (Merck, Suprap ur, 30%) and nitric acid (Merck, Suprapur, 65 %) as well as MilliQ purified water (Thermo Scientific MicroPure Ultrapure Water System, 18.2 M cm) were used. All centrifuge tubes used in this work were rinsed with MilliQ purified water and only plastic and glass were used for processing radioactive samples to avoid contact with metal contaminants. 145 The radiolabeling solution was pH adjusted wi th metal free sodium acetate (Alfa Aesar, anhydrous power, 99.997%) prepared in MilliQ purified water. 1 mg of DTPA - TOC was dissolved in ultrapure water to make a 10 mM solution. A portion of this stock solution was diluted to 1mM using ultrapure water for use in the radiolabeling experiments. Silica gel thin layer chromatography plates (Merck, TLC Silica gel 60 F 254 ) w ere used with sodium citrate dissolved in MilliQ purified water and pH adjusted with sodium hydroxide (Macron, ACS reagent grade) for the TL C mobile phase. The ICP standards were prepared from 8 Trace Cert mixed solutions for ICP from Millipore Sigma that included 67 elements altogether: 100 ppm Alkali metal mix in 2% nitric acid (Li, Na, K, Rb, and Cs), 100 ppm Alkali earth metal mix in 2% ni tric acid (Be, Mg, Ca, Sr, Ba), 100 ppm Transition metal mix 1 in 2% nitric acid (V, Cr, Mn, Fe, Co, Ni, Cu, Zn, Ag, and Cd), 100 ppm Transition metal mix 2 in 2% nitric acid and 1% hydrofluoric acid (Ti, Zr, Hf, Nb, Ta, Mo, W, and Re), 100 ppm Transition metal mix 3 in 10% hydrochloric acid (Ru, Rh, Pb, Os, Ir, Pt, and Au), 100 ppm Post - transition metal mix in 2% nitri c acid (Al, Ga, In, Tl, Pb, and Bi), 100 ppm Metalloid and non - metal mix in 2% nitric acid and 0.5% hydrofluoric acid (B, Si, P, S, Ge, As, Se, Sn, Sb, and Te), and 50 ppm Rare earth metal mix in 2% nitric acid (Sc, Y, La, Ce, Pr, Nd, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, and Lu). These standards were diluted with the appropriate concentration of Suprapur HNO 3 and HCl required for the standards to match the solution matri ces of the samples analyzed . When running the calibration and sample analyses, Suprapur 3% HNO 3 and 3% HCl were used for the HNO 3 and HCl sample matrices, respectively. Additionally, the falcon tubes used were rinsed at least tw ice with ultrapure water to remove any background stable ions before samples were added. 146 6 .2.1.2 Instruments Identification and quantification of radionuclides were performed with an HPGe Canberra BEGe Gamma - ray Detector (BE2020). The HPGe detector was cal ibrated for energy and efficiency with a 152 Eu point source at 25 and 50 cm from the detector fac e. Genie 2000 software (Mirion Technologies) was used to analyze gamma spectra. An AccuSpin Micro 17 microcentrifuge (Thermo Fisher Scientific) was used at 700 0 g/min for ease in working with the concentrated 47 Sc radiolabeling sample . A n Amersham Typhoon T M Biomolecular Imager (GE Healthcare, Chicago, USA) and phosphor imaging film s ( BAS - IP SR 2040, GE Healthcare ) were used to visualize radioactivity on TLC plat es . These images were analysed with ImageQuant TL software ( version 8.2.0.0, GE Healthcare ) to pe rform background subtractions and quantify the relative intensities observed in the image. Stable elemental analysis was performed with an Agilent 5900 SVDV IC P - OES equipped with an Advanced Valve System and a 0.5 mL sample loop. 6 .2.2 48 Ca Irradiation Th e second - generation isotope harvesting target at the A1900 beam blocker position was irradiated with a 140 MeV/nucleon 48 Ca beam (see Figures 2 .5 and 2 .7). Dur ing the experiment, the beam intensity was increased from 1 to 80 pnA over 5.1 hours for a total integrated beam intensity of 80 pnA h. The irradiation was separated into four irradiation periods of increasing beam intensity. After each of the first three periods , a water sample was withdrawn through the sample line in the aqueous chemistry box. Follo wing removal of the third water sample, a cation exchange resin bed was inserted in the system in the aqueous chemistry box and remained in the system for 16 h ours . This collection period included the 147 Figure 6 . 2 : Beam Structure and Sample Collection The measured beam current is shown as a function of time (blue line). A cation exchange resin bed was inserted bef ore the fourth irradiation period and removed 16 hours later with a wate r sample withdrawn concurrently. A second cation exchange resin bed was inserted at that time and was removed 24 hours later with another water sample. fourth irradiation period as w ell as several hours after the end of the irradiation. The first resin b ed was removed for use in offline experiments and a second cation exchange resin bed was used to continue collection of 47 Ca for the next 24 hours. Based on HPGe measurements of the fi rst resin bed and a water sample removed at the time this resin bed was removed, a total of 1.79(6) mCi of 47 Ca w ere produced during this irradiation. The beam cu rrent structure and the time points at which sampling was performed during the experiment are shown in Figure 6 . 2 . 6 .2.3 Production of 47 Ca Production rate measurements were carried out by removing 250 mL water samples from the thoroughly mixed isotope harvesting water after the first three irradiation periods. Gamma - ray spectrometry used to identi fy and quantify radionuclides in the sample. The 148 activities me asured in the 250 mL water samples were scaled up to estimate the total activity of each radionuclide in the total water volume of the system. After completion of the experiment, an accurate val ue for the total water volume was found by spiking a known vol ume and concentration of CaCl 2 solution in to the flowing - water system, allowing the water to thoroughly mix (as determined by constant conductivity readings of the system water), and withdrawing a sample of known volume. The concentration of calcium in the water sample was determined with ICP - OES, a nd the dilution of the calcium concentration was used to calculate the total water volume in the system when each of the three water samples were with drawn. An additional correction was made to the gamma - ray spectroscop ic measurements to account for the n on - standard geometry of the samples. This correction factor was found by making gamma - ra y measurements of a small volume ( i.e. , 250 µ L) which resemble d a point source and contain ed each of the radionuclides listed in Table 6 .3 (see Section 6 .3.1). The 250 µ L sample was then quantitatively transferred to 250 mL of water to replicate the same geometry as the 250 mL water samples withdrawn during the exper iment. The measurements from these two sample geometries were compared to find a geometry correction facto r for the 250 mL water samples. Approximately 5 to 20% fewer counts (depending on the gamma - ray energy and the distance from the detector) were observ ed from the 250 mL water sample compared to the smaller 250 µ L volume most likely due to more extensive at tenuation of gamma - rays in the larger water volume . Both the measurement of the water volume in the system and the geometry correction factors are dis cussed in more detail in Appendix D. Based on the geometry - corrected total activity measurement of each ra dionuclide in the water system and the beam current during each corresponding irradiation period , a production 149 rate, in terms of particles produced per incoming beam particle was found with Equation 6 .1 for each radionuclide in each irradiation period : ( 6 .1) where is the number of pro duced nuclei remaining at the end of the irradiation , is the beam current during the i th irradiation interval from to , is the decay constant of t he produced radionuclide, and is the time between the i th irradiation interval and the end of the irradiation. This segmented production equation was used to account for fluctuations in beam intensity during each irradiation period . 6 .2.4 Collection of 47 Ca from Isotope Harvesting System After the radioactive products were formed in the interior of the flowing water target, the water transport ed these products from the target to the rest of the isotope harvesting system. This system contain ed componen ts that measure d the conductivity of the water to understand the level of d issolved ions in the water and that remove d ions from the flowing water to use these products in further studies. In this experiment, cation exchange resins were used to remove 47 Ca 2+ from the water. Essentially a ll the cationic radioactive products as wel l as stable cations in the water were collected on the cation exchange resin beds. When designing the resin beds, two important functions were considered: 1) effective collection o f ions from the flowing water and 2) removal of ions from collection resin bed for use as the loading solution in further purification. To satisfy both requirements, 1.5 g of resin was used for the collection resin bed. This quantity has a large capacity t o capture the radionuclidic products even in the presence of significant st able ion contamination while being 150 small enough that a multiple resin bed volume rinse for elution purposes remains practical. The ideal solution resulting from elution has the smal lest volume with the lowest acidity possible. This allows for a dilution st ep instead of a time consuming and less practical evaporation step to prepare the load solution for the HCl gradient purification process. The cation exchange resin AG 50Wx8 in the form of large diameter resin beads (mesh size 20 - 50) was used in all the co llections described in this work. The large bead size allowed for a rapid flow of water through the resin bed without a large pressure drop across the resin bed . 1 fitting on each end of the tube. The glas s wool immobilized the resin against the flow of water and the fitting on each provided a water - tight seal between th e tube housing the resin and the - coupling bo dy, NS4D10004 and NS4D24004) were also used on each side of the resin bed in the system to allow for rapid, spill - fre e removal of the resin beds from the system (depicted in Figure 2 .13). The collection started about 4 hours into the irradiation after the withdrawal of water samples for the production rate measurements. The first cation exchange resin collect ed cations for 16 hours including 4 hours during the irradiation and 12 hours following the irradiation . A second cation exchange resin bed was then u sed for the next 24 hours to increase the total collected activity. The average flow rate of the water over the resin beds was maintained around 400 mL/min for this experiment (430 mL/min for the first resin bed, 380 mL/min for the second). The percent of the 47 Ca collected was used as a metric to understand 151 the collection efficiency with these resin beds. When a resin b ed was removed from the system during either experiment, a water sample was withdrawn from the system. Gamma - ray spectroscopy was used to m easure the activity of each radionuclide on the column and in the water samples. The activity measured in the water s ample was then scaled up to account for the entire volume of the water system. The percent collected was found using the activity collected on the cation exchange resin beds compared to the remaining activity in the water as determined with the water sampl e. Once collected, the 47 Ca was eluted from the collection resin beds with 1 or 2 M HCl. These relatively low concentrations of HCl were u sed to find the ideal solution, as discussed previously, to produce a final 47 Ca solution with as small a volume and as low an acidity level as possible. The first resin bed was rinsed with 71 mL of 1 M HCl in fractions of 5 - 10 mL followed by 18 mL of 2 M HCl in 5 mL fractions. The second resin bed from this experiment was rinsed with 40 mL of 2 M HCl in fractions of 5 m L. Each fraction and each of these resin beds before and after elution were measured with gamma - ray spectroscopy to determine the percent e lution for each radionuclide. 6 .2.5 Purification of 47 Ca The load solution for the purification of 47 Ca was prepar ed from the first 47 mL of 1 M HCl used in the elution of the first collection cation exchange resin bed. This solution was diluted to 95 m L using ultrapure water to produce a solution of 0.5 M HCl and was divided into four equal fractions. Three of these fractions were used to perform the purification process in replicate, while the fourth was left unpurified. Each fraction contained 0.28(1) mCi of 47 Ca at the time they were prepared. 152 The three replicate purifications were performed using the same column containing 2 g of AG MP - wool and a PVDF (polyvinylid each end of the tube. The cation exchange resin was preconditioned with approximately 25 mL of 0.5 M HCl before each separation. A pump speed of 1.5 mL/min was used throughout the separation . The approximately 24 mL load solution at 0.5 M HCl was added to the column followed by a 28 mL 2 M HCl rinse step. This load and rinse step w ere designed to remove 7 Be, 24 Na, 28 Mg, 42,43 K as well as any stable Fe from the column. Elution of 47 Ca was carr ied out in approximately 20 mL of 4 M HCl. Throughout the separation, all scandium isotopes loaded remained on the co lumn due to their high distribution coefficient on this resin at these conditions. After the elution step, the column was rinsed with about 50 mL of water prior to storage. The steps of this separation as well as the general behavior of the radionuclides a re illustrated in Figure 6 . 3 . An additional purification step was used for one of the purified 47 Ca solutions following the AG MP - 50/HCl method. This step involved passing the 4 M HCl solution over a 1.5 g DGA resin and using an additional 20 mL of 4 M HC l to ensure a quantitative remo val of 47 Ca from the resin. Then , 3 M HNO 3 was used to elute any stable Fe that was removed from the purified 47 Ca solution as it passed through the column. These 3 M HNO 3 samples were analyzed for their stable ion content as described in Section 6 .2.8. These separations were evaluated using the separation yield and the radionuclidic purit y of 47 Ca. The separation yield s measured the percentage of 47 Ca that was recovered in the 4 M 153 Figure 6 . 3 : Separation Scheme for the Purification of 47 Ca The color gradient of the arrows indicates the increasing molarity of acid used in the gradient elution. The placement of the radionuclides in the fi gure show where each species was at the be ginning (load solution) or at the end of the separation (rinse solutions or on the column). HCl used in the final elution step. This value was found for each separation as the ratio of the 47 Ca activity in the purified 4 M HCl elution compared to the or iginal 47 Ca activity in the loa d solution, where the activity in the purified fraction w as decay corrected to the time that the load solution was measured. The radionuclidic purity of the 47 Ca in the 4 M HCl solution was taken to be the percentage of 47 Ca activity in the final purified solution compared to the total activity of all radionuclides in the solution. The activity of 47 Sc was not included in the total activity in the solution as this activity inevitably grows in after the purification of 47 Ca. Ad ditionally, the activity of 45 Ca was included in the total activity in the solution since it follows the chemical behavior of 47 Ca in each step and decays to stable 45 Sc, and therefore, does not affect the radionuclidic purity of the final 47 Sc solution . 154 6 .2.6 Generation of 47 Sc The purified solution of 47 Ca in 4 M HCl was used as the generator solution for 47 Sc. After 2 to 4 days of decay time, 68 to 93% of the theoretical maximum 47 Sc activity w ill be generated. Therefore, the generator solution was use d approximately 3 days post - purification ( i.e. , the AG MP - 50/HCl separation or for successive generations of 47 Sc, the pseudo generator separation) . This approach provided a good balance between generati ng a sufficient 47 Sc activity for radiolabeling while not wasting time or 47 Ca by waiting until the maximum activity that would occur after approximately 5.6 days of decay. The theoretical activity of a 47 Ca/ 47 Sc generator is shown in Figure 6 . 4 for an ini tial 47 Ca activity of 0.28 mCi . A sligh tly modified version of the pseudo generator separation procedure outlined by Domnanich et al (2017) was used in this work to purify 47 Sc. [10] This p rocedure involves two Figure 6 . 4 : 47 Ca/ 47 Sc Generator Activity The calculated 47 Sc activity generated with 0.28 mCi of purified 47 Ca is shown as a function of time. The two arrows indicate the 47 Sc activity at 3 and 5.6 days of decay time. 155 main steps using two DGA extraction chromatography columns to first purify the 47 Sc and then concentrate the 47 Sc activity in a sm a ll volume (Figure 6 . 5 ). Both columns were preconditioned with about 10 mL of 4 M HCl. In t he first step, the 4 M HCl generator solution was loaded onto a 50 - HCl, Ca 2+ p asses through the column easily while Sc 3+ is extracted from solution and retained on the column. A small 4 mL rinse of 4 M HCl was then us ed to rinse the load solution vessel and the column to ensure all the 47 Ca activity passed through the column. Then, 7 mL of 0.1 M HCl w as used to remove 47 Sc from the column. The second step in the pseudo generator process involved first re - acidifying t he 47 Sc solution from the first DGA column to 3 M HCl through the addition of about 3 mL of 9.46 M HCl. This loading solution acidity ensured the retention of 47 Sc on the second DGA resin column his solution was passed over the column at a flow rate of 0.3 mL/min, allowing any Figure 6 . 5 : Modified Pseudo Generator Procedure The procedure proceeds from left to right starting with loading the genera tor solution on the first DGA column and ending with eluting 47 Sc from the second DGA column. 156 residual 47 Ca to pass through and the 47 Sc to be retained in a tight band on the 20 - 25 mg of DGA resin. Both the low flow rate and the small mass of resin were selected to reduce the dispersion of the 47 Sc and assist in th e concentration step. A rinse step of 7 mL 3 M HNO 3 was added to the original pseudo generator procedure to remove an y residual Fe that was not removed during the purification of 47 Ca. This step was added due to low radiolabeling efficiencies observed afte r multiple uses of the original pseudo generator procedure. After this rinse step, a 7 mL 4 M HCl rinse step w as used to retain 47 Sc on the column while switching the acid matrix on the column. With a flow rate of 0.5 mL/min, 47 Sc was eluted for the column in 0.05 M HCl and collected in the following fractions: 1) 1 - 2 drops, 2) 700 L, 3) 500 L, 4) 500 L, 5) 2 mL. Thes e fractions were designed to catch the largest portion of the activity in the second 700 L fraction for use in radiolabeling. As with th e purification of 47 Ca, this separation procedure was characterized by finding the separation yield and radiopurity o f the final purified 47 Sc sample. These values were found by using an HPGe de tector to find the activity in each fraction and on both DGA 1 and 2 before and after each loading step. The separation yield from DGA 1 was found by comparing the activity on the column after loading and after elution. The yield for DGA 2 was found by com paring the activity in the load solution with that in the seco nd 47 Sc fraction. A total separation yield was found as the product of these two yields from DGA 1 and 2. Due to the long half - live of 46 Sc, it is both important and more difficult to measure t he activity of 46 Sc in the final purified sample s . Therefore, a long measurement of each radiolabeling solution was taken between 24 and 3 5 days after this purification was perfor med. For samples where 46 Sc w as not detected, the limit of detection for this radionuclide was found 157 by taking the square root of the sum o f counts over 7.5 keV centered around 889 keV, which corresponds to a characteristic gamma ray energy for this radion uclide. The activity found with this method is given as the upper limit of th e activity of 46 Sc in the sample and is used to find a lower l imit for the radionuclidic purity for 47 Sc in the radiolabeling solutions. These values are given for each radiolabel ing solution in Table 6 .4 in Section 6.3.4 . 6 .2.7 Radiolabeling DTPA - TOC with 47 Sc The secon d 47 Sc fraction from the second DGA column in the pseudo generator procedure contained the majority of the 47 Sc activity in a volume of approximately 7 00 L. Approximately 50 to 100 L w as remainder wa s used as the radiolabeling solution each time the pseudo generator procedure was carried out. Activity measurements were taken of the final solution (app roximately 650 L) with the HPGe detector to determine the radionuclidic purity of the 47 Sc and the sp ecific activity of the final radiolabeled complex. The first step in preparing the radiolabeling solution was to pH adjust the solution to a level at wh ich both the Sc 3+ and the DTPA - TOC peptide w ould be in a suitable chemical form. Since scandium tends to hydrolyze at pH levels approaching neutral conditions, an acidic pH ensures that scandium is dissolved in solution as Sc 3+ . However , peptides like TOC can degrade under strongly acidic conditions through cleavage of the peptide bond between the amino ac ids by acid hydrolysis. A mildly acidic solution around pH 3 to 3.5 can be used to maintain scandium solubility and prevent significant degradation of the peptide. With a portion of the solution set aside, pH adjustment tests were carried out for each generator. With the second DGA column of the generator using 3 M HCl immediately before elution of 47 Sc in 0.05 M HCl, there is a 158 potential for residual acid in the column to change the acidity of this elution step. Additionally, there is potential variability in the extent of this change depending on the condition of the column when this elution step starts and how much volume is collected in the first 47 Sc fr action. For these reasons, these pH adjustment tests are performed for each replicate of the generator procedure. These adjustment tests involved combining different volume ratios of the radiolabeling solution and the 0.5 M sodium acetate buffer that was u sed to adjust the pH. These solutions were vortexed and spun down in the microcentrifuge before testing the pH with pH strip paper. The ratio that produced a pH of 3 to 3.5 was used to pH adjust the radiolabeling solution. In almost all cases, the ratio of the radiolabeling solution to the buffer used to achieve this pH was 4:1. The pH of this solution was also checked with pH strip paper before the addition of the pep tide. The volume of 1 mM DTPA - TOC required to achieve 1.5 MBq/nmol was added to each radio labeling solution. The final solution was shaken by hand and spun down for 10 seconds with a microcentrifuge. The exact formulations used for each radiolabeling solut ion are given in Table 6 .1. The radiolabeling solutions were heated gently at about 80 - 85 ° C for 10 minutes to encourage chelation between the 47 Sc 3+ and DTPA - TOC in solution. Once the solution had cooled, it was again shaken by hand and spun down with a mi crocentrifuge. For radiolabeling solution 1, the concentrated 47 Sc fraction was split in t wo, resulting in two solutions described in two rows in the table. For radiolabeling solution 2 to 4, the sodium acetate b uffer and the first, smaller addition of DTPA - TOC listed in the table were added as a first radiolabeling attempt. The second row in t he table represents an addition of more DTPA - 159 Table 6 . 1 : Radiolabeling Con ditions Radiolabeling Solution Generator Number Activity Used (MBq) Buffer Added ( Total Peptide Content (nmol) Specific Activity ( MBq/nmol) 1 1 0.640(5) 62.5 0.5 1.28(1) 0.636(5) 62.5 2 0.318(2) 2 2 1.74 (1) 150 1.2 1.34(1) 1.65(1) - 1.7 0.895(7) 3 3 1.33(1) 150 1.2 1.02(1) 1.09(1) - 2.4 0.418(4) 4 5 3.60(3) 120 1.2 3.00 (2) 3.29(3) - 3.2 1.03(8) 5 6 1.36(1) 95 1.2 1.1 4(1) TOC followed by another 10 - minute heating step in a second attempt at raising the radiolabeling yield. The total amount of DTPA - TOC adde d to the radiolabeling solution at this point is given in the second row for solutions 2 to 4. No addi tional buff er was added in this second attempt as indicated by the dash in the second row for t hese solutions. Thin Layer Chromatography (TLC) was used a s a quality control test to determine the radiolabeling efficiency, defined as the percent of 47 Sc rad iolabeled c ompared to the total activity in the radiolabeling solution. The method used silica gel TLC plates on an aluminum backing as the stationary pha se and a solution of 0.1 M sodium citrate in 0.2 M HCl to achieve a pH of 4.7 for the mobile phase. Wi th these co nditions, any free [ 47 Sc]Sc would migrate up the TLC plate with an R f value of 0. 5 a nd any [ 47 Sc]Sc - DTPA - TOC complex would remain at the origin where originally spotted (Figure 6 . 6 ). Dilutions of both the free [ 47 Sc]Sc control solution ( i.e. , th e 50 - 100 L set aside from the second fraction collected from DGA 2) and the radiolabeling 160 Figure 6 . 6 : TLC Quality Control Test Schematic description of t wo TLC lanes used for each test: the control with [ 47 Sc]Sc - D TP A (left) and the radiolabeling solution with [ 47 Sc]Sc - DTPA - TOC and/or [ 47 Sc] Sc - DTPA (right). solution were prepared for use in the TLC quality control tests . Both solutions were diluted in a ratio of 9:1 (water to [ 47 Sc]Sc solutions) with ultrapure wat er and a final concentration of 30 M DTPA. These dilutions were prepared for two reasons: to reduce the activity spotted on the TLC plates in 1 L of each solu tion to avoid overexposure of the phosphor imaging films used to quantify the radiolabeling effici ency and to a dd DTPA to both solu tions to chelate free [ 47 Sc]Sc. The DTPA chelated form of [ 47 47 Sc]Sc more centralized and uniform and therefore, easier to quantify with the phosphor imager without changing the R f val ue of the spo t. After performin g the TLC test, the plates were dried and placed in plastic bag to contain the dispersible radioactivity. The plates were then laid on a phosphor imaging film in a dark environment for about 3 minutes. The TLC plates were r emoved, and the film was read by the 161 phosphor imager. ImageQuant TL software was used to define regions of inter est at both R f = 0. 5 and 0 for both the control and the test lane. A background correction was made for each of these regions of interest and th e percent activity in the two reg ions of interest was found for each lane. 6 .2.8 Stable Elemental Analysis Stab le elemental analysis was performed with a broad spectrum ICP - OES method. As used previously for the first 48 Ca irradiation (see Chapter 4 Secti ons 4 .2.6 and 4 .3.5), a semi - quan tification method was used to identify and quantify any stable ions present in the samples analyzed. This method was chosen for the wide range of elements and concentrations that can be simultaneously identified and quantif ied in a small sample volume. In the present experiment, the IntelliQuant broad spectrum analysis was used on a different instrument and with a more careful calibration than previously used. While this method is typically used for identification and semi - q uantification of stable ions, the instrument can be carefully calibrated before use each time to produce reliabl e quantification of stable ions. The careful calibration of the instrument and uncertainty quantification for the measurements is a significant difference between the use of a s emi - quantitative ICP - OES method used previously in the first 48 Ca irradiation a nd the present investigation. The accurate quantification was supported by linear readings of standards from 5 ppb to 5 ppm for most elements a nd with a few quality control che ck standards used before and after all samples were analyzed. The relative devi ation of the standards compared to the line of best fit and the deviation of the quality control check standards from their true concentrations were added in quadrature to find an uncertainty for the quantification of each element. This resulted 162 in about 4 - 15% uncertainty in the quantification of all elements identified in the samples. The calibration was performed for different sample matrices si nce the presence of high acid con centrations can change concentration readings due to changes in the physical ch aracteristics of the solution. Since most samples fell into one of four sample matrix categories, calibrations were performed for standards in a 3 M HNO 3 , 1 M HCl, 2 M HCl, and 4 M HCl matrix. For each matrix, standards were used from 5 ppb to 5 ppm with a pure acid such as Ca, Fe, and Sc. The identical i nstrument settings w ere used for both the calibration and the sample analysis and are given in Appendix C. 6 .3 Results and Discussion 6 .3.1 Production of Radionuclides with 48 Ca Beam Production rates from the 140 MeV/nucleon 48 Ca beam irradiation of a f lo wing - water target were measured f or all radionuclides with an activity above the limit of detection with HPGe gamma - ray measurements at 10 to 20 min after the irradiation period and are given in Table 6.2 production rates are given in nuclei produced per i ncoming 48 Ca beam particle. The se values are also converted to a percent to represent the percent conver sion of the primary beam to each radionuclide by atom. Additionally, these measured production rates were used to extrapolate the activity that would be produced in the isotope harves ti ng system if it were to be used to collect left - over beam for a typical 48 Ca primary beam experiment at the NSCL. The activities are those immediately after the irradiation period and after a 24 - hour waiting period in which the radionuclides decay before e xperimenters access the activity. A correction was 163 Table 6 . 2 : Production Rates of Radionuclides in Flowing - W ater Target with 140 MeV/nucleon 48 Ca Beam Production rates reported as radionuclide produced per i nc omi ng 48 Ca beam particle and as percent conversion of the beam to each radionuclide. Extrapolated activities based on the measured production rates are g iven for isotope harvesting for a typical 48 Ca primary beam experiment at the NSCL ( i.e. , 4.5E11 pps at 14 0 MeV/nucleon for 120 h) and after a 24 - hour cool down period following this irradiation. When a dash is given, it indicates that the activity level h as dropped below 10 Bq. Nuclide Production Rate Percent Conversion Act ivity (GBq) After Irradiati on After 24 - hour cool down Na - 24 8. 2 ( 1 )E - 4 0.08 2 ( 1 ) 0.3 67(7 ) 0.12 1 (2) Mg - 27 3. 07 ( 2 )E - 4 0.03 07 ( 2 ) 0.1 38 ( 1 ) - Mg - 28 1. 9 ( 1 )E - 4 0.01 9 ( 1 ) 0.08 1 ( 5 ) 0.037( 2 ) S - 38 1.2 8 ( 7 )E - 4 0.012 8 ( 7 ) 0.05 7 (3) 1. 6 (1)E - 4 Cl - 34m 3.9(1)E - 5 0.0039( 1 ) 0.017 5 (6) - Cl - 38 1.16( 4 )E - 3 0.116( 4 ) 0.5 7 ( 5 ) - Cl - 39 6. 5 ( 5 )E - 4 0.06 5 ( 5 ) 0.29( 2 ) - K - 42 3. 34 (7)E - 3 0.3 34 (7) 1.5 0 (3) 0. 391 (8) K - 43 4. 1 ( 1 )E - 3 0.4 1 ( 1 ) 1.8 1 ( 6 ) 0.8 6 ( 3 ) K - 44 2.6( 5 )E - 3 0.26( 5 ) 1.2(2) - K - 45 1.6( 2 )E - 3 0.16( 2 ) 0.7 1 ( 9 ) - Ca - 47 1.6 7 ( 4 )E - 2 1.6 7 ( 4 ) 4.0 2 ( 9 ) 3.4 5 ( 8 ) Sc - 44m 4 . 69 (8)E - 4 0. 0 4 69 (8) 0.16 0 ( 3 ) 0.12 0 (2) Sc - 46 5.2( 6 )E - 3 0.52( 6 ) 0.09 (1) 0.09( 1 ) Sc - 47 4 .62 ( 5 )E - 3 0.4 62 ( 5 ) 2.8 ( 4 ) 3. 0 ( 6 ) Sc - 48 2. 17 ( 3 )E - 3 0.2 17 ( 3 ) 0.8 3 ( 1 ) 0.5 67 ( 8 ) applied to the measured production rates for 47 Sc and 38 C l to accou nt for contributions fr om their co - produced parents. The reported production rates for 47 Sc and 38 Cl in Table 6 . 2 reflect the rate of direct production for these radionuclides. However, the activities extrapolated in T able 6 . 2 account for both di rect produ ction and decay contr ib utions from their parent radionuclides. 164 In a previous isotope harvesting experiment with a 48 Ca primary beam described in Chapter 4 , the production rate for 47 Ca was found to be 0.020(4) with a large systematic error . [41] The measured rate of 0.016 7 ( 4 ) in this experiment agrees well with t he previous value with a much smaller uncertainty. The production rates were measured with two different methods, giving cre di bility to the accuracy of the result . Th e production rate for 47 Ca is the highest measured in this irradiation and is high for charged particle reactions in general. Given this production rate, a typic al experiment at th e NSCL using a 48 Ca primary beam c ou ld produce 4.0 2 ( 9 ) GBq of 47 Ca for the generation of 47 Sc. The 48 Ca primary beam anticipated for isotope harvesting at FRIB is higher in energy and more than 400 time s more intens e than the available b eam at the NSCL. Th is should result in a higher produ ct ion rate and a higher number of possible nuclear interactions, respectively. 6 .3.2 Collection of 47 Ca from Isotope Harvesting System Collection of 47 Ca from the 50 L isotope harvesting water system wa s achieved with a c ation exchange resin with high eff ic acy. During the present experiment, 92(5)% of the 77(3) MBq of 47 Ca produced was collected on a single 1.5 g cation exchange resin bed with a water flow of 430 mL/min over the resin bed for 16 hours. T he remaining 47 Ca i n the water was removed with a sec on d cation exchange resin bed of the same size over 24 hours at a similar flow rate of 380 mL/min. Overall, 101(5)% of the 47 Ca was removed from the water with the 47 Ca activity remaining in a 500 mL sample of water below the limit of detection with an HPGe gamma - ray detector after the second cation exchange resin wa s remov ed. 165 Figure 6 . 7 : Collection Resin Bed Elution Pr ofiles Elution profile for the first (a) and second (b) cation exchange resin beds used for collection of 47 Ca . 166 The 47 Ca was effectively removed from the cation exchange resin beds for further processing using 1 or 2 M HCl. The elution profiles for all radionuclides on the cation exchange resin bed are given for these two elutions in Figure 6 . 7 . For bo th elutio ns, Ca, Na, K, and Mg isotopes elute in a sharp peak at the beginning of the elution, with their peaks reaching a maximum value earlier and occurring in a smaller volume for the 2 M HCl elution as expected. 47 mL of 1 M HCl eluted 83(5)% of the 47 Ca on the first cation exchange resin bed used in the collection step. For the second collection cation exchange resin bed, 21 mL of 2 M HCl was used to elute 82(4)% of the 47 Ca adsorbed to the resin. This earlier peaking in the elution curve results in a si milar e lution percentage in less than half the volume for the 2 M HCl. This would result in a smaller dilution factor to reach the same acidity level for the purification load solution using the 2 M HCl elution compared to the 1 M HCl elution. All Sc is ot opes el ute to a much lower extent throughout the elution and 40 to 60% of the activities of these isotopes remain on the column after the elution. The Sc isotope elution curve with the most reliable trend for the 1 M HCl elution is 47 Sc as it is present on the co lumn with the highest activity. The other Sc isotopes are present at a lower activity than other radionuclides in the elution fractions at the time that the measurements were made, making reliable measurements of the activities of the Sc isotopes di fficult in the presence of the relatively large background created by other radionuclides. Since the elution of the second column was performed a few days after that of the first, the activity of 44m,46,48 Sc in each frac tion and on the resin bed are more a ccurate ly measured with a significant decay of other shorter - lived radionuclides present on the resin bed. This can be clearly observed as the elution profiles of all Sc isotopes from the second cation exchange resin bed are almost 167 indistinguishable with i n the e rror bars on the points. These profiles also are similar in character to that observed for 47 Sc from the first resin bed. 6 .3.3 Purification of 47 Ca The AG MP - 50/HCl method was successfully repeated three times w ith the eluate from the first coll ec tion re sin bed used. HPGe gamma - ray measurements of each fraction for one of these replicates were used to produce the elution curve in Figure 6 . 8 . These separations demonstrated the reproducibility and efficacy of this purification method with an averag e separat ion yield of 97.5(3)% at a radionuclidic purity of 99.8(4)% immediately following the separation. During the time that 47 Sc was generated after this purification step, the main impurities in the 47 Ca samples (i.e. , 42,43 K) had decayed significantl y, result ing in a radionuclidic purity of 100.0(4)% at the time that the generated 47 Sc was separated from its parent. Figure 6 . 8 : Elution Profile for Purification of 47 Ca 168 6 .3.4 Generation of 47 Sc The sep ar ation yields from DGA 1, DGA 2, and the complete pseudo generator procedure for each replicate of the generator procedure are given in Table 6 . 3 . The separation yield from DGA 1 is high at >97% of the 47 Sc in the 7 mL el ution step except for that for gen er ator 4 (discussed below). The yield is lower for DGA 2 due to the limited volume required for radiolabeling to maximize the interactions between 47 Sc and DTPA - TOC. This separation yield was found only for the activity el uted in the 700 L 47 Sc fra ction f ro m DGA 2, which is shown as the highlighted portion of the elution profile in Figure 6 . 9 . The total separation yield 47 Sc in the generator solution w as purified in t he final 70 0 L sa mp le. This process also significantly concentrated the 47 Sc activity from the generator solution which contained about 20 mL for generator 1 to 3, 50 mL for generator 4, 150 mL for generator 5, and 160 mL for generator 6. Figure 6 . 9 : Elution Profile of 47 Sc from DGA 2 graph. 169 Table 6 . 3 : Generator 1 to 6 Results The activities of 47 Ca and 47 Sc in the generator were measured be fore the start of the pseudo generator process. All activities used to calculate the separation yields were decay corrected to the end of the generator procedu re. Generator # Activity of 47 Ca (MBq) Activity of 47 Sc (MBq) Separation Yield from DGA 1 (%) Sep aration Yield from DGA 2 (%) Total Separation Yield (%) 1 4.12(1) 2.10(2) 99(1) 95.3(1) 94.5(2) 2 3.40(2) 2.38(2) 97.7(7) 84.5(2) 82.5(3) 3 3.32(1) 2.31(2) 98(1) 76.0(9) 75 (1) 4 3.18(1) 3.34(3) 94(1) 38.6(4) 35.4(6) 5 4.98(8) 4.95(4) 99(1) 79.3(9) 78( 1) 6 2.03(4) 1.66(1) 98(1) 91(1) 89(1) 47 Sc fraction was determined to be quite high for all generators except for generator 4 with no activity other than 47 Sc being detected in the gamma - ray spectra taken ri ght after the generator procedure and about one month later. The limit of detection for 4 6 Sc in the spectra from one month after the generator procedure was fo und for all gene rators except generator 4 and used to find a lower limit on the radiopurity of th e 47 Sc radiolabeling solutions. This limit for the activity of 46 Sc and the radionuclidic purity of 47 Sc are both given in Table 6 . 4 for each radiolabeling sol ution. Generat or 4 purified the fourth fraction from collection resin bed 1 that was not purifi ed using the AG MP - 50/HCl method. This fraction contained about 25 mL of 0.5 M HCl and th e mixture of radionuclides that eluted from the collection cation exch ange resin bed: 7 Be, 24 Na, 28 Mg, 43 K, 47 Ca, 44m, 46,47,48 Sc. This fraction was purified using the pseudo generator solution to demonstrate that all non - scandium radionuclides left in the sample have the same elution behavior as 47 Ca. Even with these radion uclides present, they do not interfere with the 170 purification of 47 Sc through the pseudo generator as shown in Figure 6 . 10 . The generated 47 Sc isolated in this example was accompanied by the all the sufficiently long - lived scandium isotopes produced in the fragmentation re actions in the isotope harvesting target. As these scandium isotopes were produce d in a water medium and circulated in the system in a near neutral pH, they were most lik ely hydrolyzed to a oxide form with low solubility ( e.g. HScO 2 or Sc 2 O 3 ) and filtered out of the water on the glass wool in the cation exchange resin bed. Therefore, t he scandium species present in this separation scheme were not all Sc 3+ in solution. This led to unpredicted elution behavior and the poor yields shown in Tabl e 6 . 3 . Figure 6 . 10 : Elution from DGA 1 in Gener ator 4 The x - axis gives the phase of the separation while the y - axis gives the percent of each radionuclide that had eluted. The graph demonstrates that 7 Be, 2 4 Na, 28 Mg, 43 K, and 47 Ca elute with almost identical profiles, leaving only the scandium isotopes in the final eluted solution for concentration with DGA 2. 171 Using the elution behavior shown with generator 4, an alternative purification procedure can be prop osed. To remove the scandium isotopes initially, DGA 1 from the pseudo generator procedure coul d be used as a purification step after elution of the radionuclides from the collection column. As demonstrated by the elution profile in Figure 6 . 10 , this step would leave isot opes of Be, Na, Mg, K, and Ca in the generator solution while removing all scan di um isot opes. This semi - purified generator solution would be sufficiently prepared to produce a sample of generated 47 Sc with high radionuclidic purity after a wait period of 3 to 5 days and using the pseudo generator procedure. A further discussion of th e merits and flaws of this separation procedure is given in Chapter 9 . 6 .3.5 Radiolabeling DTPA - TOC with 47 Sc The details for each radiolabeling solution check ed with the TLC quality control are listed in Table 6 . 4 . Modest specific activities of 0.3 - 3 MB q of 47 Sc /nmol of DTPA - TOC were attempted since the level of 47 Sc activity available was low. The TLC quality control results for Table 6 . 4 : Radiolabeling Results Solution Generator Number Activity of 47 Sc ( MB q) Activity of 46 Sc (Bq) Radio - nuclidic Purity (%) Specific Activity (MBq/nmol) Percent 47 Sc Radiolabeled Percent 47 Sc Free 1 1 0.640(5) <104 >99.9826 1.28(1) 11 89 1 1 0.636(5) 0.318(2) 88 12 2 2 1.74(1) <102 >99.9936 1.34( 1) 26 74 2 2 1. 65(1) 0 .8 95(7) 38 62 3 3 1.33(1) <92 >99.9912 1.02(1) 43 57 3 3 1.09(1) 0.418(4) 35 65 4 5 3.60(3) <127 >99.9990 3.00(2) 0 100 4 5 3.29(3) 1.03(8) 0 100 5 6 1.36(1) <102 >99.9931 1.14(1) 100 0 172 radiolabeling solutions 1 - 3 (fr om generators 1 - 3, respectiv el y) demonstrate that a low level of radiolabeling occurred. Specifically , the 10 to 40% of 47 Sc that successfully radiolabeled is not high enough to be considered successful and indicates that there is a metal contaminant inter fering with the chelation be tw een 47 Sc and DTPA - TOC. Radiolabeling solution 4 resulted from combining t he generator solutions produced from generator 1 - 4 and running the generator procedure a fifth time (generator 5). No matter the level of DTPA - TOC used, no radiolabeling was observe d with this solution. Again, this result indicates that there was a metalli c contaminant in the radiolabeling solution that limited the success of the radiolabeling. Since the generator solution resulting from generator 4 was no t purified with the AG MP - 50 /H Cl purification step, it is possible the a higher amount of a metallic io n remained in the generator solution even after processing it with the pseudo generator procedure. The addition of this generator solution most likely in troduced enough of this cont am inant to drop the radiolabeling success rate from the low level observed with radiolabeling solutions 1 - 3 to 0% with the combined solutions. The only notable change that occurred between generator 5 and 6 (producing radiolab eling solutions 4 and 5, res pe ctively) was the addition of the 3 M HNO 3 rinse step in the generator pro cedure. With this change, the radiolabeling yield increased dramatically to essentially 100%. Since the 3 M HNO 3 rinse step was designed to remove metall ic impurities su ch as Fe, th is strongly indicates that a small amount of ionic impurity in the generato r solution was concentrated in the radiolabeling solution and interfered with previous radiolabeling attempts. An example phosphor image from the TLC qua lity control tes ts for radio la beling solutions 3, 5, and 6 are shown in Figure 6 .1 1 . In the first image , the single spot in 173 Figure 6 . 11 : Phosphor Images for Thin Layer Chromatography Quality Control Tests Phosphor image for TLC from (a ) radiolabeling solution 4 with no radi olabeling ( i.e. , 0% yield), (b) radiolabeling solution 3 with partial radiolabeling f or both specific activities used ( i.e. , 43% yield for 1 MBq/nmol in the middle lane and 35% yield for 0.4 MBq/nmol in the right mo st lane), (c) and radiolabeling solution 5 with complete radiolabeling ( i.e. , 100% yield). In all three images, the leftmost s pot is the control spot of [ 47 Sc]Sc - DTPA at an R f value of about 0. 5 . the radiolabeling solution test lane (right) at the same R f v alue as the control [ 47 Sc]Sc - DTPA solution (left) indicates no radiolab eling occurred i n radiolabeling solution 4. The spots at the origin and at about R f = 0. 5 for radiolabe ling solution 3 at both specific activities used (middle and right lane) demonst ra tes the partial radiolabeling, while the single spot at the origin for radiolabeling so lution 6 shows complete radiolabeling. 6 .3.6 Stable Elemental Analysis T he results of th e stable elemental analysis on the eluted fractions from the cation exchange re sin beds are given in Tables 6 . 5 and 6 . 6 . The stable ions in these fractions demonstrate that mg quantities of Al and Mn, hundreds of µ g of Ca and Ni, and lower levels of Co, Cr , Cu, Fe, Mg, and Zn were present in the water system and collected on the ca ti on exchange resins during the course of this irradiation. Additionally, Si was identified in these samples. Due to non - linear responses from Si at this concentr a tion range, a qu antification of this element was not made. It is likely that this ion content r esulted from stainless steel ( e.g. , 174 Fe, Cr, Mn, Ni, Si) and aluminum alloy ( e.g. , common aluminum which is 3003 aluminum alloy contains Al, Mn, Fe, Si, Cu, Zn i n decreasing mass percent) components in the system that were in contact with the water. The s amples tested from the purification of 47 Ca and 47 Sc are given in Table 6 . 7 . After the collected radionuclides were eluted from the first cation exchange resin b ed, the solution was diluted and separated into four equal volume fractions as explained in S ec tion 6 .2.5. The stable content of these fractions is shown in Table 6 .7 grouped 47 1 to 3 which were purified through the AG MP50 / HCl method have very low stable ion masses except for Ca 2+ . Fraction 4 which was not purified a fter elution from the cation exchange resin bed contains significant levels of several elements, especially Al, Mn, and Ni. This comparison demonstrates the abi l ity of the AG MP - 50/HCl purification method to remove stable impurities from the final purifi ed 47 Ca product. It is also clear that the DGA resin step used from the first fraction from the cation exchange resin bed is not needed since the AG MP - 50/HCl met h od provides suff icient purification and there is not significant difference between the ICP - O ES results for Fraction 1 before and after the DGA resin step. Since the generator solutions resulting from all four fractions of purified 47 Ca ( i.e. , Generator s olution 1 - 4) wer e combined after each fraction was processed with the generator procedure onc e, the total stable content of Fractions 1 - 4 are essentially equal to that measured in Generator solution 7. Characterization of this generator solution demonstra t es that the pres ence of significant levels of Al, Co, Cu, Mg, Mn, and Ni did not interfere wi th radiolabeling since they pass through DGA 1 into the new generator solution with Ca. Notably, the Cr, Fe, and Zn content in the unpurified Fraction 4 are absen t from Generator solution 7. Since no 175 significant stable ion content was observed in the 3 M H NO 3 rinse step added to Generator 6 (shown as Generator 6 DGA 2 Fr 2 in Table 6 .6), it is most likely that the Cr, Fe, and Zn content were removed from the genera t or solution when generator #4 or 5 were performed. These stable ions eluting with the 47 Sc fr ac tion in Generator 5 would explain why the radiolabeling solution from Generator 5 resulted in a radiolabeling yield of 0%. With these ions removed from the gene r ator solution, t he radiolabeling solution produced with Generator 6 would have contained lowe r levels of competing ions, resulting in a 100% radiolabeling yield. Potentially, if the 3 M HNO 3 rinse is used from the first generator purification, interfering ions such as Cr, Fe, and Zn could be removed from the generated 47 Sc. 176 Table 6 . 5 : ICP - OES Analysis of Eluate from Cation Exchange Resin Bed 1 Fractions 6 - 13 from the elution of the first cation exchange re sin bed were ana lyzed with ICP - OES. These were the fractions that were not used for the next pu rification step. Collection Fraction Mass (µg) Al Ca Co Cr Cu Fe Mg Mn Ni Zn Cation Exchange Resin Bed 1 Fr 6 1 M HCl 170(20) 22(2) 1.6(2) 0.60(5) 10(1) 2.4 (2) 1.5(1) 260(2 0) 100(10) 2.3(2) Fr 7 1 M HCl 180(20) 24(2) 0.68(6) 7.4(8) 2.2(2) 1.7(1) 27 0(20) 100(10) 2.1(2) Fr 8 1 M HCl 160(20) 20(2) 0.52(5) 5.9(6) 2.0(2) 1.2(1) 200(10) 76(9) 1.7(1) Fr 9 1 M HCl 230(30) 30(2) 2.4(2) 0.99(9) 13(1) 4.4(3) 3.3(2) 390(30) 1 50(20) 3.7(3) Fr 10 1 M HCl 130(20) 19(2) 1.2(1) 0.51(4) 6.1(6) 2.3(2) 1.9(1) 20 0(10) 73(9) 1.9(2) Fr 11 2 M HCl 370(20) 34(3) 0.81(3) 1.6(2) 9.0(6) 4.6(4) 2.8(2) 250(20) 96(7) 2.5(2) Fr 12 2 M HCl 250(20) 23(2) 1.1(1) 5.5(4) 3.1( 3) 1.8(1) 150(10 ) 63(5) 1.5(1) Fr 13 2 M HCl 230(10) 17(2) 0.92(9) 3.7(3) 2.5(2) 1.4(1) 120(10) 43 (3) 0.98(9) Total: 1.7(1) x 10 3 200(10) 6.0(6) 7.0(4) 60(4) 23(1) 15.7(7) 1.83(9) x 10 3 700(60) 17(1) 177 Table 6 . 6 : ICP - O ES Analysis of Eluate from Cation Exchange Resin Bed 2 T he second cation exchange resin bed was eluted in eight 5 mL fractions of 2 M HCl. Collection Fraction Mass (µg) Al Ca Co Cr Cu Fe Mg Mn Ni Zn Cation Exchange Resin Bed 2 Fr 1 2.6 x 10 3 150(10) 7 0(8) 9.2(9) 180(10) 27(2) 38 2.3(2) x 10 3 2.4(2) x 10 3 4 5( 4) Fr 2 2.0 x 10 3 110(10) 6.5(7) 103(7) 18(2) 22(2) 1.4(1) x 10 3 1.4(1) x 10 3 27(2) Fr 3 910 56(5) 18(2) 3.5(4) 40(3) 8.4(7) 9.5(9) 610(50) 570(40) 9.8(9) Fr 4 640 38(4) 11(1) 2.3(2) 20(1) 5. 5(5) 5.8(5) 380(30) 340(30) 5.1(5) Fr 5 520 25(2) 6.9( 8) 1.7(2) 11.6(8) 4.3(4) 3.8(3) 270(20) 220(60) 3.1(3) Fr 6 380 18(2) 4.3(5) 1.3(1) 6.6(5) 2.9(2) 2.6(2) 180(10) 140(10) 1.8(2) Fr 7 280 12(1) 2.8(3) 1.0(1) 3.9(3) 2.4(2) 1.5(1) 11 6(9) 88(6) 1.2(1 ) Fr 8 280 9.9(9) 2.5(3) 1.2(1) 3.3(2) 2.6(2) 1.5(1) 1 06 (8) 80(6) 1.2(1) Total: 7.6(2) x 10 3 200(10) 6.0(6) 7.0(4) 60(4) 23(1) 15.7(7) 1.83(9) x 10 3 700(60) 17(1) 178 Table 6 . 7 : ICP - OES Analy sis of Samples f rom the Purification of 47 Ca and 47 Sc The first 25 mL of 1 M HCl used to elute 47 Ca from the first cation exchange resin bed were diluted to 100 mL of 0.5 M HCl. This solution was divided into four fractions described in the table as Fr 1 - 4 . The results fr om fraction 1 are give n nin g before and were purified with the AG MP - 50/HCl method and fraction 4 a ilu tion step. The total mass for each element across Fr 1 after, Fr 2 purified, Fr 3 purified, and Fr 4 not purified is given at the bottom row of the table. Sample Mass (µg) Al Ca C o Cr Cu Fe Mg Mn Ni Zn Generator 6 DGA 2 Generator 7 1.0(1 ) x 10 3 790(50) 24(2) 150(10) 31(2) 3.0(2) X 10 3 1.28(8) X 10 3 47 Ca Purification Fr 1 before 220(10) 2.0(1) 4.5(3) Fr 1 after 210(10) 3.7(2) 4.3(3) Fr 2 purified 210(10) 0.26( 2) 0.74(4) 4.8(3) Fr 3 purified 200(10) 0.44 (2) 3.3(2) Fr 4 not purified 1.0(1) x 10 3 210(10) 28(2) 5.6(4) 170(10) 20(2) 25(2) 3.0(2) x 10 3 1.36(8) x 10 3 46(3) Total 1.0(1) x 10 3 830(30) 28(2) 5.8(4) 170(10) 20(2) 29(2) 3.0 (2) x 10 3 1.36(8 ) x 10 3 46(3) 179 6 .4 Conclusion A n irradiation of the is oto pe harvesting system with a 140 MeV/nucleon 80 pnA 48 Ca beam was performed, produc ing approximately 1. 8 mCi of 47 Ca . The 47 Ca activity was effectively collected from the 50 L water s ystem and purifi ed through an AG MP - 50/HCl gradient separation method. Sta ble elemental analysis demonstrated that significant levels of several elements such as Al, Mn, Ca, and Ni were present in the water and collected on the cation exchange resin bed. A majority of the stable ions were removed from the purified 47 Ca throug h t he AG MP - 50/HCl method. The purified 47 Ca was used as a generator solution for the generation of 47 Sc. Repeated use of the gen erator procedure likely resulted in the removal of low l evels of Cr, Fe, and Zn which may have interfered with radiolabeling. T he pseudo generator procedure presented previously was used with this isotope harvested sample of 47 Ca/ 47 Sc to produce concentrat ed samples of 47 Sc with radionuclidic purities of >99.99 %. These samples were used to troubleshoot the generator and radiolabel ing procedures to radiolabel DTPA - TOC with a radiochemical yield of 100%. The experiment described here allowed for testing and optimizing each step in the process of isotope harvestin g from producing 47 Ca in the isotope harvesting system to using purifie d 4 7 Sc generated from 47 Ca for radiolabeling experiments. The high quality [ 47 Sc]Sc - DTPA - TOC produced with the harvested sample o f 47 Ca demonstrated the feasibility of using radionuclid es produced thro ugh isotope harvesting for nuclear medicine research. T he procedures demonstrated with a low - level activity sample of 47 Ca in this experiment can be applied to experiments with h igh er integrated beam currents , producing a higher activity sa mple of 47 Ca . 180 Chapter 7 : 48 Ca Beam Experiment 3: High Activity 47 Ca/ 47 Sc Genera tor and 47 Sc Radiolabeling with Isotope Harvested 47 Ca A full power 48 Ca beam experiment was performed at the NSCL with the isotope harvesting target in the beam blocker position. Thi s experiment pro duced approximately 10 mCi of 47 Ca that was effectively co llected from the 50 L water system. Th is radionuclide was purified wi th an AGMP - 50/HCl gradient purification method to produce a 47 Ca/ 47 Sc generator with high radionuclidic purity. T he generated 47 S c was purified in three successive replicates and was u sed to radiolabel DTPA - TOC with a preclinically useful specific activity of 6 MBq/nmol. 7 .1 Introduction Following the optimization of procedures for the purification of generated 47 S c and radiolabel ing DTPA - TOC with 47 Sc described in Chapter 6 , another hig her integrated beam power 48 Ca irradiation was performed at the NSCL . The irradiation produced approximately 10 mCi of 47 Ca which was used to produce a 47 Ca/ 47 Sc generator . The purif ication of 47 Sc f rom the generator for use in radiolabeling experiments wa s performed three times in succession to demonstrate the reproducibility of producing [ 47 Sc]Sc - DTPA - TOC. This experiment was designed to produce a high quality [ 47 Sc]Sc - DTPA - TOC prod uct at a high en ough activity to use in preclinical targeted internal r adi otherapy studies. Additionally, this study demonstrate d the feasibility of using isotope harvesting as a production technique for radionuclides with medical applications. 181 7 .2 Materia ls and Methods 7 .2.1 Materials The materials and instruments used in th is experiment are described in Chapter 6 Section 6 .2.1. 7 . 2.2 48 Ca Irradiation The present experiment started with a 2 - hour irradiation segment at 60 pnA beam as a warm - up period to mo nitor operations of the system under irradiation and then ran at 80 pnA fo r 7.7 hours. T hus, the experiment extended over 11.7 hou rs with a total integrated beam intensity of 703 pnA - h. There were n o intentional interruptions of the irradiation of the targ et. The gaps in the beam current shown in Figure 7 .1 were necessary for be am tuning before increasing the beam intensity to 80 pnA and due to issues with the accelerated beam upstream of the target. Figure 7 . 1 : Experiment al Beam Structure The green vertical line labeled on resin was in place at the start of the irradiation. This resin bed collected almost all of the activity and was removed for processin g 24 hours after the end of the irradiation. 182 7 .2.3 Collection of 47 Ca f rom Isotope Harvesting System A single cation exchan ge resin bed was used throughout the nearly 12 - hour irradiation and for 24 hours after the irradiation to collect as much 47 Ca activi ty as possible i n one batch on the resin bed. The water flow rate over the resin bed for collection was higher than that us ed in the previous experiment (see Chapter 6 Section 6 .2.4) at 570 mL/min. A second resin bed was used to collect the remainder of th e activity follo wing the removal of the first resin bed. Based on the c oll ection resin bed elution results from the previou s experiment, the first cation exchange resin bed used in this experiment was rinsed with 25 mL of 2 M HCl and the eluent collected i n a single fract ion. Both the collection efficiency of 47 Ca from the s yst em water and the elution efficiency of 47 Ca from the cation exchange resin bed were measured by gamma - ray spectroscopy of the two resin beds and the eluted fraction. Measurements of the activity on the first cation exchange resin bed were made post - elut ion on the same day and several days after the eluti on due to the high activity of the scandium isotopes which in large part remained on the resin. The activity in the eluted fraction w as estimated by measuring the activity in a 0.1 mL aliquot which had a low enough activity to be quantified at a reasonable distance from the detector face. An elution efficiency was calculated by comparing the activity measured in the eluted fraction to t he total activit y in the eluted fraction and remaining on the resin bed . T o find the collection efficiency of the system a separate measurement of the second cation exchange resin bed was made. Based on the 100% collection efficiencies reported in the prev ious experiment (Chapter 6 Section 6 .3.3) and an identical collection p rot ocol used in the present experiment , a 100% colle ction efficiency between these two resin beds was assumed. The collection e fficiency for only the 183 first cation exchange resin bed was found based on this assumed total collection efficiency and the gamma - ray spectroscopy measurements of the eluted fraction, the first cation exchange resin bed post - elution, and the second cation exchange resin bed. 7 .2.4 Purification of 47 Ca The AG MP - 50/HCl separatio n procedure for the purification of 47 Ca from Chapter 6 Se ction 6 .2.5 was slight ly a dapted for this experiment. Instead of performing r eplicate purifications with the collection resin bed eluate, the entire acid rinse from this resin bed wa s purified in a single batch . To accommodate this change and avoid the eva poration step, a dilution of the collection resin bed eluate was performed to reduce the acidity and prepare a n appropriate load solution for the purification. Multiple load solutio n dilution condi tions were tested with stable K + and Ca 2+ ions: 500 mL of 0.1 M HCl, 250 mL of 0.2 M HCl, and 167 mL of 0.3 M HCl. Each of these stable - ion load solutions w as followed by a small rinse solution used to rinse the load phase container, the t ubing, and the c olumn (20 mL of 0.1 M HCl, 5 mL of 0.2 M HCl, and 5 mL of 0.3 M HCl, respectively). The 250 mL 0.2 M HCl load solution was also used with either a 40 or 100 mL 0.2 M HCl rinse solution following the load step. Th ese larger 0.2 M HCl rinse s teps w ere used w hen testing the separation to capitalize on the differe nce between the distribution coefficient for K + and Ca 2+ on AG MP - 50. Since 0.2 M HCl was high enough acidity to decrease the distribution coefficient for K + (K d = 165) while that for C a 2+ remains high (K d = 3.7 x 10 3 ) . [80] The final purification procedure used in this experiment was based on these stable element tests and previous results presented in Cha pter 4 Section 4 .3.3 and Chapter 6 Section 184 6 .3.3. The 25 mL of 2 M HCl used to elute 47 Ca from the collection resin bed was diluted to 250 mL 0.2 M HCl. An additional 40 mL of 0.2 M HCl was used as the first rinse step followed by a 26 mL 2 M HCl rinse. Th e final elution of 47 Ca was performed with 32 mL of 4 M HCl. 7 .2.5 Generation of 47 Sc The purification of 47 Sc from the 47 Ca/ 47 Sc generator was performed in a similar manner to tha t described in C hapter 6 Section 6 .2.6. Figure 7 .2 lays out the steps i n t his purification method. The main difference between the present method and the method used for the final generator solution in the previous irradiation (Generator 6 in Chapter 5) is the timing of t he 3 M HNO 3 rinse step in the process. This rinse step was used following the 4 M HCl load phase on the first DGA resin rather than following the 3 M HCl load phase for DGA 2 as done in the previous experiment . Since the first DGA resin is about 3 times la rger in mass than DGA 2 and the volume used to elute 47 Sc from DGA 1 is le ss critical than that used for DGA 2, this change allowed for a higher rinse volume with a higher flow rate. The large volume of 30 mL of 3 M HNO 3 was used to ensure the removal of a ny stable Fe from the column, particularly because the mas s of Fe present in the system was not known when this decision was made during the experiment. Following this rinse, a 4 M HCl rinse of 10 mL was used to change the acid matrix in th e column back to HCl before the elution of 47 Sc. These rinse phases wer e c arried out at 1. 2 mL/min as with the other rinse steps for DGA 1. All other rinse steps, flow rates, and fraction collections were carried out as described in Chapter 6 Section 6 .2.6 . 185 Figure 7 . 2 : Schemat ic Description of the 47 Ca/ 47 Sc Generator Procedure 7 .2.6 Radiolabeling DTPA - TOC with 47 Sc The second 47 Sc fraction taken from DGA 2 in the generator procedure was the radiolabeling s olution. This so lution was prepared as described in Chapter 6 Section 6 .2. 7 by using approximately 50 L of this fraction for pH adjustment tests, adding the proper ratio of sodium acetate buffer following these test s , adding DTPA - TOC to reach the desired specific activit y, and heating the final solution to encourage interact ion between 47 Sc and DTPA - TOC. Gamma - ray spectroscopy measurements were made of the radiolabeling solution before and after the 50 L sample was removed . The 47 Sc solution to sodium ace tate buffer rati os used for pH adjustment of radiolabeling sol utions 1 - 3 w ere 4:1, 5:1, and 3:1, respectively. The activity of 47 Sc, volume of buffer, and nanomoles of DTPA - TOC used in each radiolabeling solution are given in Table 7 .1. Each combination re sulted in a spec ific activity just over 6 MBq/nmol. 186 Table 7 . 1 : Radiolabeling Conditions Radiolabeling Solution Activity of 47 Sc (mCi) Activity of 47 Sc (MBq) Buffer Added ( µ L) Peptide Added (nmol) Specific A ctivity (MBq/nmo l) 1 1.871(3) 69.2(1) 178 11.2 6.18(1) 2 1.033(2) 38. 21( 6) 139 6.0 6.41(1) 3 0.316(2) 11.70(7) 235 1.9 6.16(4) To check the radiolabeling yield, thin layer chromatography using a sodium citrate mobile phase was carried out as describe d in Chapter 6 S ection 6 .2.7 and depicted in Figure 6 .6. Due to the hig her activity used in these radiolabeling experiments, a two - stage dilution was used to produce a low er activity solution necessary fo r phosphor imaging. In the first stage, a 9:1 water to 47 Sc solution ratio w as used. The second step used a 50:1 dilution o f w ater to 47 Sc solution. This final solution also contained about 30 nM of DTPA to chelate any free 47 Sc. Both the free 47 Sc solution ( i.e. , the 50 uL of the second 47 Sc fraction from DGA 2) and the r 47 in these dilutions. The TLC plates were exposed to the phosphor imaging films for five minutes, and then, the resulting images were analyzed with ImageQuant TL software as described in Chapter 6 Secti o n 6 .2.7. 7 .2.7 Stable Elemental Analysis ICP - OES anal ysi s was performed on the 3 M HNO 3 rinse fractions used in the generator procedures, the final Generator 4 solution, and all unused fractions from the two AG MP - 50/HCl separations. The unused samples f or the separations included the 0.2 M HCl fractions fro m t he first separation and the 0.2 M, 2 M, and last 4 M HCl fractions from the second separation. The same instrument and materials listed in Chapter 6 Section 6 .2.1 and the same setti ngs, 187 standards, calibrations, and uncertainty quantification described in Chapter 6 Section 6 .2.8 were used in this analysis. All uncertainties for the observed elements were in the range of 6 to 15%. 7 .3 Results and Discussion 7 .3.1 Collection of 47 Ca fr om Isotope Harv e sting System Between the irradiation and the cool down per iod of 24 hours , 95(3)% of the 362(9) MBq of 47 Ca produced was removed from the water with a single 1.5 g cation exchange resin bed. After removal from the system, the resin bed was processed with 25 mL of 2 M HCl to elute 84(3)% of the collected 47 Ca. Th is higher acidity rinse was used in the elution from the collection resin bed, as less than half the volume is required to elute a similar percent when compared with 1 M HCl. This wa s demonstrated w ith a 1 M HCl elution from resin bed 1 and a 2 M HCl el uti on from resin bed 2 in the irradiation previously described in Chapter 6 Section 6 .3.2. Table 7 . 2 : Stable K + /Ca 2+ Test Separations Lo ad Phase Rinse S eparation Yield for Ca 2+ (%) Purity of Ca 2+ (%) 500 mL 0. 1 M HCl 20 mL 0.1 M HCl 97.5 99.9 250 mL of 0.2 M HCl 5 mL 0.2 M HCl 98.9 99.7 40 mL 0.2 M HCl 99.1 99.9 100 mL 0.2 M HCl 98.8 99.7 167 mL of 0.3 M HCl 5 mL 0.3 M HCl 96.2 99.8 188 7 .3.2 Purific ation of 47 Ca An AG MP - 50/HCl purification procedure wa s c arried out based on the three successful purification replicates from the previous experiment (Chapter 6 , Section 6 .3.3). The procedure was adjusted so that the entire eluate from th e co llection res in bed was diluted and purified in one batch to avoid a n e vaporation step. Separation tests with stable Ca 2+ and K + ions were performed to optimize the dilution factor used for the load phase. The various load solutions and resulting Ca 2+ s epar ation yields and purities for these test separations are given in T abl e 7 .2. While all of the conditions tested resulted in high yields and purities for the Ca 2+ fraction, the 250 mL 0.2 M HCl load solution with an additional 40 mL 0.2 M HCl rinse prod uced the best re sults. This load solution and rinse phase relied on the di fference in distribution coefficients of Ca and K while controlling zone broadening of the ions on the column due to large volumes and thus minimize d the overlap between the two elut ion peaks. The e lution profile for the test separation under these cond iti ons is shown in Figure 7 .3. Figure 7 . 3 : Stable Ca 2+ /K + Separation This procedure used a 250 mL 0.2 M HCl load phase and a 40 mL 0.2 M HCl rinse pha se followed by a 2 M HCl rinse and finally the 4 M HCl elu tion phase. 189 Figure 7 . 4 : Separation Elution Profiles for AG MP - 50/HCl Separations The first (a) and second (b) 47 Ca purification from the present experiment . The dashed lines indicate a change in the mob ile phase and the text above each section provides the mobile phase composition. indicate the percent of each radionuclide that remained on the column through the entire separation. The uncertainties given in the data p oints represent the counting statistics. T he optimized 250 mL 0.2 M HCl load phase was used to purify 47 Ca produced in this irradiation. Unfortunately, the active 47 Ca produced during this experiment had a diff erent elution pr ofile from that fo r in the stable element separation tests with Ca and K as can be observed by comparing the elution profiles in Figure 7 .3 and Figur e 7 .4a. The first separation performed with in the present experiment eluted 49.0(6)% of th e 47 Ca in the 27 .2 m L 2 M HCl 190 r ins e step , while only 0.9% of the stable Ca was eluted in the 26 mL 2 M HCl rinse step in the separation procedure with stable ions shown in Figure 7 .3. Additionally, th e separation performed with stable ions showed that Ca c ontinued to elut e slowly in add iti onal volumes of 2 M HCl with only 5.3% o f the stable Ca eluted in 33 mL of 2 M HCl. Due to this shift in the elution profile, only 50.9(4)% of the 47 Ca in the load solution for this separation was purified with a radionucl idic purity of > 99.9%. A second se paration was performed with the 2 M HCl rinse step solution that contained the remaining unpurified 4 9% of the 47 Ca. The elution profile in Figure 7 .4b demonstrates a successful purification of 47 Ca was obtained with a 26. 6 m L 2 M HCL rin se step. This s epa ration resulted in a high separation yie ld of 99.1(7)% with 99.9(8)% radionuclidic purity. The two pu rified 47 Ca fractions from these separations were combined and resulted in a total separation yield of 97(1)% with a radi onuclidic purity of 99.9(6)%. T he total recovery yield from production to purification of 47 Ca was 78(3)% for this experiment. 7 .3.3 Generation of 47 Sc The 47 Sc activities as well as the separation yield for 47 Sc wi th DGA 1 and DGA 2 in the generator procedure are given in Table 7 .3. T he yields for 47 Sc from DGA 1 were relative ly Table 7 . 3 : Generator 1 - 3 Results Generator # Activity of 47 Sc Separation Yield from DGA 1 (%) Separation Yield f rom DGA 2 (%) Total Separation Yield (%) 1 2.76 (2) 94.9(8) 72.4(1) 68.6(6) 2 1.36(1) 94.6(8) 87.1(5) 82.3(9) 3 0.661(5) 99.7(9) 53.2(5) 53.1(7) 191 consistent at about 95% or higher. For DGA 2, the yields for 47 Sc in fraction 2 were lower, as expected due to the limited volume collected in fraction 2. The elution profile for DGA 2 of generator 1 and 3 are sho wn in Figure 7 . 5 . Th e profile is characterized by a sharp peak at low volumes followed by a tailing elution behavior for severa l milliliters. T he separation yield for 47 Sc from DGA 2 was substantial ly lower for generator 3 of this experiment compared to th e separation yield from DGA 2 found for either generator 1 or 2 from t his irradiation or those presented in Chapter 6 Section 6 .3.4. The low yi eld for the final generator in this experiment is possi bly due to the use of a high molarity HCl load solution followed immediately by a low molarity HCl rinse to elute the 47 Sc from DGA 2 . A high yield in a small volume low molarity HCl ph ase relies on th e low distribution coefficient for Sc on DGA resin unde r t h ese conditions. However, any residual acidity or droplets from the high molarity load solution can mix with the elution phas e and raise the actual acidity of this phase as 47 Sc is e luted. It is lik ely that this contributed to the low yield for generato r 3 . This explanation is supported by the amount of Figure 7 . 5 : Elution P rofile for DGA 2 in Generator s 1 and 3 The dashed orange/bl ue area indicates 72.4% (a) and 53.1% (b) of the 47 Sc activity was collec ted in a fraction of about 700 L in Generator 1 and 3, respectively. The lower yi eld for Generator 3 resulted from the wider elution peak than that from Generator 1. 192 Table 7 . 4 : Radiolabeling Results Radiol abe ling Solution Activity of 47 Sc (MBq) Activity of 46 Sc (kBq) Radionuc lidic Purity Percent 47 Sc Radiolabe led Percent 47 Sc Free 1 69.2(1) 2.3(2) >99.99 96.5 3.5 2 38.21(6) < 0.5 >99.99 99.1 0.9 3 11.70(7) < 0.3 >99.99 100 0 sodium acetate required to pH adjust the final radiolabeling solution. A hig her ratio of 47 Sc solution to buffer ( i.e. , a lower amoun t of buffer adde d) would indicate a lower molarity of HCl in the radiolabeling solution. This ratio was found to be roughly correlated with the 47 Sc s epa rat ion yield for DGA 2 with ratios of 4:1, 5:1 , and 3:1 used for generators 1 , 2, and 3, respectively. Generator 3 in t his experiment required the lowest ratio to reach a pH of 3 to 3.5 indicating that it contained the highest molarity of HCl. 7 .3.4 Ra dio lab eling DTPA - TOC with 47 Sc The radiolabeling solutions were analyzed with TLC to determine the radiochemical yield for [ 47 Sc] - Sc - DTPA - TOC and with gamma - ray spectroscopic measurements to determine the radionuclidic purity of the 47 Sc. Each o f the radi ola bel ing experiments had a yield of > 95% as shown in Table 7 .4 and a specific activity of at least 6 MBq/nmol. The phosphor images of the TLC tests shown in Fig ure 7 .6 demonstrate this high radiochemical yield. At the time of radiolabeling, only 47 Sc was obs erv ed through gamma - ray measurements. Additional measurements 5 months later allowed for quantification of an 46 Sc present in th e radiolabeling solution. This radionuclide was found only in radiolabeling solution 1 at a low level as shown in Table 7 .4 . The li mi t of detection for this radionuclide by its characteristic gamma - ray energy of 193 Figure 7 . 6 : Phosphor Images of Thin Layer Chromatography Quality Control Tests Pho sphor images of radiolabeling soluti on 1 - 3 in a - c. the single spot in at the origin in the right lane for each TLC test indicates hig h radiolabeling yields for each solution. 889 keV are also given in Table 7 . 4 for radi olabeling soluti on s 2 and 3. The radionuclidic purity for each solution wa s at least 99.99% with only a low level of 46 Sc present i n the first solution. T he high radiochemical yield and radionuclidic purity as well as the moderate specific activity demonst rate that the fi nal radiolabeled [ 47 Sc] - Sc - DTPA - TOC product was of suff ici ent quality for preclinical studies. 7 .3.5 Stable Elementa l Analysis The results of the stable elemental analysis of several samples from this irradiation are given in Table 7 . 5 . N one of the 0.2 M HCl samples from both AG MP - 50/HCl separations ha d sign ifi can t concentrations of stable ions. The 2 M HCl fractions as well as the last 4 M HCl fraction from the second separation contained the masses shown in the lower half of Table 7.5 . The 2 M HCl samp les contained mg quantities of Al, Mn, and Ni as well as te ns of µ g quantities of Co, Cr, Cu, Fe, Mg, and Zn. The las t 4 M HCl fraction was not used for the generation of 47 Sc due to a low concentration of 47 Ca. This sample with low levels of Ca, Mg, and M n is 194 representative of the stable ion content of the 4 M HC l f ractions that were used in the following chemistry steps. This can also be seen in the stable ion content of the Generator 4 solution with hundred s of µ g of Ca and lower levels o f Mg and Mn. T he 3 M HNO 3 rinse step used in Generators 1 - 3 are given in the upper half of Table 7.4 . The rinse step for the first generator was measured twice and gave consistent but low - level readings for Fe, Mg, and Zn. These stable ions that were rem oved in this rin se step are similar to those that were removed through r epe ati ng the generator procedure in Chapter 6 Section 6 .3.6 . In both experiments, after the removal of these ions, a high radiolabeling yield was achiev ed. Therefore, future experiment s should be desi gned to ensure the removal of Fe, Zn, and Cr from the 47 Ca/ 47 S c generator. 195 Table 7 . 5 : Stable Element Results from Generators 1 - 4 and AG MP - 50 #2 Separation Group Fraction Mass ( µ g) Al Ca Co Cr Cu Fe Mg M n Ni Zn Generator #1 DGA 1 Fr 2 3 M HNO 3 1.6(2) 0. 114 (6) 3.7(7) #2 DGA 1 Fr 2 3 M HNO 3 #3 DGA 1 Fr 2 3 M HNO 3 #4 solution 4 M HCl 470(30) 11.2(7) 16(1) AG MP - 50 #2 Fr 7 2 M HCl 1.5(1) x 10 3 0.35(3) 23(3) 2.2(2) 82 (6) 67(6) 40(3) 2.4(2) x 10 3 1.24(9) x 10 3 45(4) Fr 8 2 M HC l 3.4(2) x 10 3 0.67(6) 9(1) 8.4(8) 21(1) 64(5) 12(1) 1.8(1) x 10 3 440(30) 6.2(6) Fr 9 2 M HCl 680(40) 2.3(2) 3.4(3) 0.57(4) 4.5(4) 0.46(4) 48(4) 12.1(9) Fr 15 4 M HCl 1.8( 1) 0.16(1) 0 .31(1) 196 7 .4 Conclusion Th e present experiment demonst rat ed the feasibility of isotope harvesting to produce 47 Sc - based radiopharmaceuticals for preclinical studies. An experiment was conducted with the flowing - water aqueous target as the A1900 beam bloc ker at the full NSCL 48 Ca beam power . The 47 Ca produced was ef ficiently collected on a single collection resin bed, eluted from this resin bed, and purified with an AG MP - 50/HCl method. Evidence of stable ions was discovered through the une xpected elution behavior during the purification of 4 7 Ca. As a result, a la rge 3 M HNO 3 rinse step with the first DGA resin of the generator procedure was used to purify 47 Sc of metal impurities. ICP - OES measured significant levels of Al and Mn among sever al other lower l evel stable ions from the isotope har vesting water syste m . The 3 M HNO 3 rinse step used in the generator procedure was also shown to remove Fe and Zn which may be impurities that interfere with radiolabeling efficiencies. The generator proc edure was used t o produce a sample of high purity 47 S c which was then us ed to radiolabel DTPA - TOC with a high radiochemical yield and at the preclinically relevant specific activity of 6 MBq/nmol. Th us, a series of procedures ha s been successfully optimize d to produce hig h quality [ 47 Sc] - Sc - DTPA - TOC for furt her research. 197 Chapter 8 : Measurement of the Three Most Intense Gamma Rays Following the Decay of 47 Ca A sample of 47 Ca produced by isotope harvesting at the National Superconducting Cyclotron Labor atory was used t o measure branching ratios of 7.17(5)%, 7.11(5)%, and 75. 0(5 )% for the 489.2, 807.9, and 1297.1 keV characteristic gamma rays, respectively , following the beta - decay of 47 Ca . Based on these updated branching ratios, the ground state to grou nd state 47 Ca to 47 Sc beta decay branching ratio was inferred to be 17.7( 5)% an d the ground state to 1297.1 keV excited state as 82.2(5)%. These values represent a greatly increased preci sion for all five branching ratios compared to the currently accepted values [51] . The measure ments presented here were made relative to the ingrown 47 Sc daughter in a 47 Ca sample and the well - established 159.4 keV gamma - ray branching ratio and the half - life of 47 Sc [81 83 ] . These measurements were supported by verifying that the half - lives measured with the characteristic gamm a - ray peaks over multiple spectra for both 47 Ca and 47 Sc were consist ent with previou sly reported values. Additionally, the half - lives of both 47 Ca and 47 Sc were independently measured with Liquid Scintillation Counting to re - verify the previously reported values used in this study to support the updated gamma - ray branching ratio values. 8 .1 Introduction One way to produce high specific activity sa mpl es of 47 Sc is through the production and subsequent decay of its parent radionuclide, 47 Ca [10] . When producing and processing 47 Ca, the branchi ng ratios of the most intense characteristic gamma rays are generally used to quantify 47 Ca with gamma - ray spectroscopy using High Purity Germanium (HPGe) detectors . 198 The present evaluat ed values for these gamma - ray branching ratios as well as the two main 47 Ca to 47 Sc beta decay branching ratios differ with a larger reported uncertainty from the only published measurements which were made in the 1960s [51,84] . For the quantification o f 47 Ca, the pres ent 19% uncertainty in the evaluated branching ratio of the 1297.1 keV gamma ray and the present 20% uncertainty for both the 489.2 and 807.9 keV gamma rays lead to similarly large uncertainties in the absolute activity of 47 Ca [51] . The change and uncertainty in t hese evaluated v alue s prompted remeasurement of the three main gamma rays in the decay of 47 Ca to 47 Sc as part of the present work . T he branching ratios of the characteristic gamma rays in the decay of 47 Ca can be measured by using the parent - daughter rel ationship betwee n 47 Ca and 47 Sc. In a sample of these radionuclides in which radioactiv e equilibrium has not yet been reached, the in - growth of 47 Sc relative to the decay of 47 Ca can be used to quantify the number of 47 Ca nuclei present and thereby, measur e the branching ratio s of characteristic gamma rays of 47 Ca decay . T hese updated gamma - ray branching ratio values allow for a more precise value to be found for the ground state to ground state beta decay branch and for the ground state to 1297 keV excited state beta bran ch in 47 Sc. In this work, the branching ratios for the three most intense characteristic gamma rays in the decay of 47 Ca were remeasured and used to find updated values for the two main 47 Ca beta decay branching ratios. All branching ratio values reported here have an increased precision compared to th e currently accepted values . Additionally, the half - lives of 47 Ca and 47 Sc have been remeasured using both HPGe gamma - ray spectroscopy and Liquid Scintillation Counting (LSC) and have been fo und to agree wit h thos e previously reported. 199 8 .2 Methods 8 .2.1 Production of 47 Ca The 47 Ca used in this measurement was produced with a 140 MeV/nucleon 48 Ca 20+ beam at the National Superconducting Cyclotron Laboratory (NSCL) as described in Chapter 7 . Th is ion beam was i mplan ted at an average intensity of 72 pnA for 9.8 hours (see Chapter 7 Section 7 .2.1) in a flowing - water target in which 47 Ca was produced through nuclear reactions between 48 Ca and the 16,18 O and 1 H nuclei in the water molecules. The flo wing - water targe t and water system used were described in detail in Chapter 2 Sections 2 .2.2 and 2 .3.3. The product radionuclides formed in the target were transported through the flow of water from the target to a w ater chemistry system [36] and c atio nic radionu clides, such as 47 Ca, were collected on a cation exchange resin bed (1.5 g resin, AG50W - X8, mesh size 20 - 50, BioRad) . The resin bed was removed from the system after 36 hours to allow for maximum collection of 47 Ca from the water and for short l ived radion uclides to decay. The process of implant ing the 48 Ca beam in the target, flowing water through the system, and collecting cations from the water is depicted schematically in Figure 8 .1. Figure 8 . 1 : Schematic of Main Components of Water System The 48 Ca beam is implanted in the flowing - water target, producing 47 Ca that is transported to a cation exchange resin bed where it is adsorbed. The 47 Ca is removed from the syste m on this resin bed and proces sed for further use. 200 8 .2.2 Purification of 47 Ca The purification of 47 Ca that is produced in and collected from the water system was described in Chapter 7 Section 7 .2.4. In brief, the 47 Ca was removed from the cation excha nge resin bed in 2 M HCl. This s olution was diluted to 0.2 M HCl to serve as the load solution for purification on a 2 g column of AG MP - 50 ( AG MP - 50, 100 - 200 mesh size, BioRad ). Following the load solution, rinse steps of 0.2 M HCl and 2 M HCl were used t o remove other long - lived radi o n uclides that were produced in the water system an d collected on the cation exchange resin bed in addition to 47 Ca (i.e., 7 Be, 24 Na, 28 Mg, 42,43 K). Finally, 4 M HCl was used to elute purified 47 Ca. T he first step in the pseudo generator for this parent - d a u ghter pair described in Chapter 6 Section 6.2.6 w as used t o produce both a sample of 47 Ca out of equilibrium with its daughter 47 Sc and a purified 47 Sc sample . [10] . About 10 mL of the 4 M HCl solution containing 47 Ca was passed over 67 mg of DGA extraction chromatography resin - tetra - n - octyldiglycolamide, normal resin, particle size 50 - TrisKem International) in a 1 mL ISOLUTE filtration column. The 47 Sc that was generated in this solution was adsorbed on the column and the 47 Ca parent remained in solution, resu lting in a 10 mL solution of 320 kBq (8.6 uCi) of 47 Ca. This solution was used to produce samples for LSC and HPGe measurements. About 2 mL of additional 4 M HCl was used rinse the column and remove any residual 47 Ca from the resin. Then, two 1 mL fraction s of 0.1 M HCl solution were used to remove the purified 47 Sc adsorbed on the column. The second of these 0.1 M HCl fractions, containing 250 kBq (6.7 uCi) of 47 Sc, was used to produce a sample for LSC. 201 8 .2.3 LSC M easurements Liquid scintillation countin g was used to verify the half - lives of both 47 Ca and 47 Sc since these values are crucial to the method used in this work to measure the gamma - ray branching ratios for 47 Ca. These measurements were performed with a Perkin Elmer Tri - Carb 4910 TR for three 47 C a samples at different concentrations and a 47 Sc sample. Each of the four sample s consisted of 10 mL of scintillation cocktail (Optiphase HiSafe 3, Perkin Elmer) and a small spike of activity. The three 47 Ca samples (samples 1 to 3) were prepared with 100 , spike s of the 4 M HCl solution after it was pas sed over the DGA resin, producing samples with approximately 3.2, 1.6, and 0.32 kBq of 47 Ca, respectively. This range of activities was used to help ensure that spectra were collected with suff i cient count ing statistics without saturating the detector. Due to the smaller volume and higher activity of the purified 47 Sc, a 20 times dilution was first made of the 47 approximately 0.5 kBq (10 nCi) of 47 Sc was added to 10 mL of scintillation cocktail. These samples were measured for 30 minutes each several times a day over the course of 21.7 days. A background vial containing 10 mL of scintillation cocktail was measured before and after th e four samples were measured at each time point. Additional mea surements were made at 62, 93, and 140 days after the first measurement period to observe any long - lived activity in the samples. The spectra were analyzed by summing the count ing rate in two d i fferent channel windows: 0 to 2000 and 400 to 1000. The larger 0 to 2 000 range was used for the purified 47 Sc as no major contaminants were observed in th is sample. The smaller 400 to 1000 window was selected for the 47 Ca samples since 47 Sc was generated i n these samples over time . This range was chosen because it was verified that only background count rates were 202 detected for the purified 47 Sc sample in the 400 - 1000 window ( i.e. , signals from the decay of 47 Sc were only observed below channel 400) . Theref o re, the count rate in thi s window should have decreased with the half - life of 47 Ca over the 21.7 day counting time frame since only the decay of 47 Ca contributed to signals in this restricted channel range. Each count rate measured for a sample over a cha n nel range was background corrected using the average measurement over the same range for the two background measurements made each time the set of samples was run. The error for each measurement of a sample was found by propagating the square root of the s um of the counts in the c hannel window for the sample and the background average used to background correct the sample count rate . Origin Pro 9 software was used to fit the set of values obtain ed for both 47 Ca and 47 Sc decay with user defined functions. F or the 47 Ca samples, the background corrected count rates measured over channels 400 to 1000 were fit ted with a simple exponential function and the uncertainties for each rate used as an instrumental weight : ( 8 . 1) where is the total background corrected count ing rate measured in the 400 to 1000 channel window at time from the first spectrum collected, is a fitted variable for the total background corrected count ing rate measured in the 400 to 1000 channel window a t time , and is the fitted decay constant for 47 Ca. No background variables were included as the longest time point spectra produced count ing rates equivalent to the background in the 400 - 1000 channe l window indicating there was no detectable long - l ived contaminant in the 47 Ca samples in this window . 203 For the 47 Sc sample, the background corrected counting rates for the 0 to 2000 channel window was fit ted with a two - part exponential function with the uncertainty for each rate used as an instrumental weight : ( 8 . 2) where is the total background corrected count ing rate measured in the 0 to 2000 channel window at time from the first spectrum collected; and are fitted variables for the tot al background corrected count rate measured in the 0 to 2000 channel window at time for 47 Sc and a longer - lived component, respectively; and and are the fitted decay constants for 47 Sc and a longer - lived compone nt, respectively. 8 .2.4 HP Ge Gamma - Ray Spectroscopic M easurements The remaining 4 M HCl solution of 47 Ca that was purified with the DGA resin (300 kBq or 8.5 uCi of 47 Ca) was measured in a 25 mL plastic scintillation vial with an HPGe detector (Canberra BEGe Gamma - ray Detector, BE 2 020) several times each day over the course of 21.9 days. Spectra were recorded for 30 minutes at each time point with the sample 25 cm from the face of the detector. The sample was not moved over the entire time period that spe ctra were collected to pres e rve the exact geometric relationship between the sample and the detector face. A few spectra were taken at 43 and 45 days from the first spectrum collected to ensure there were no long - lived contaminant radionuclides in the samp le and that the observed ga m ma rays followed the decay of 47 Ca at long time points. The energy scale of the HPGe detector was previously with a 152 Eu point source. Following the 47 Ca measurement s , a separate 152 Eu source suspended in epoxy (Calibrate d, NIST Traceable source; 1 g/cm 3 epoxy; 20 mL 204 fill volume) in an identical scintillation vial as the sample was measured was used to calibrate the efficiency of the detector at 25 cm from the detector face. The determination of the gamma - ray branching ratios relied on the ingrowth of 47 Sc in this 47 Ca sample, and required accurate values for the half - life for 47 Sc, a characteristic gamma - ray branching ratio for 47 Sc (i.e., the branching ratio of 68.3(4)% fo r the 159.4 keV gamma ray), and the half - life of 47 Ca. With this parent - daughter relationship and a gamma - ray spectrum of the sample, one of the Bateman equation s can be used to qu antify 47 Ca in the solution at t = 0, which corresponds to the time of the s eparation of 47 Ca and 47 Sc on DGA resin. The activity of 47 Ca at t = 0 can then be corrected to the time that the gamma - ray spectrum was taken. Altogether, the use of the Bateman e quation and the decay correction can be summarized as ( 8 . 3) where and are the activity of 47 Ca and 47 Sc at time t, respectively; and are the decay const ants of 47 Ca and 47 Sc, respectively; and time is measured from the point at which 47 Ca and 47 Sc were separated (meaning ). The gamma - ray spectra also provide the count ing rate for the three highest intensity characteristic gamma rays for 47 Ca d ecay . These count ing rates can be adjusted by the efficiency of the detector at each energy to find the emission rate from the sample. The branching ratio for each characteristic gam ma ray can then be found by taking the ratio of the emission rate for each gamma ray to the activity of 47 Ca for each spectrum. 205 Additionally, th e integrated number of events in each peak from the three most intense 47 Ca gamma rays and the 159.4 keV gamma ray f rom 47 Sc were plotted over time to verify the half - life with which t h ese each peak decay s . The uncertainty in each data point was from counting statistics ( i.e. , the square root of the integrated peak sums) and was used as an instrumental weight for each data point to find the decay rate of each gamma ray peak. For the 47 C a peaks, a simple exponential, like that shown in Equation 8 . 1, was used to find the half - life for each of the three gamma rays. Th e Bateman equation in Equation 8.3 was u sed in the following form to fit the integrated events for the 159.4 keV gamma ray pe a k with the initial 47 Sc activity fixed at zero and t = 0 corresponding to the time at which the 47 Ca/ 47 Sc separation was performed: . (8.4) The half - lives of 47 Ca and 47 Sc in addition to the initial activity of 47 Ca were allowed to vary to find an independent measurement of the half - life of 47 Sc. Verifying that each peak sum decayed according to the evaluated half - life co nfirmed that no gamma rays from othe r background or contaminant radionuclides contributed significantly to the characteristic gamma rays of interest. 8 .2.5 Error Budget For the total count ing rate obtaine d for each LSC measurement and the integrated peak sum for each HPGe measurement, the u n certainty in the se measurement s was dominated by counting statistics (i.e., the square root of the number of counts.) For the activity of 47 Sc that was found from the 159.4 keV peak and used to find the half - life of 47 Ca and 47 Sc with the 206 Bateman equation , only the uncertainty from the counting statistics was used as the uncertainty in each activity point for in the fitting algorithm. Additional uncertainties that were common to all of the 47 Sc activity d ata points were added after the half - lives were fitt e d . The uncertainty in the 159.4 keV branching ratio and detection efficiency were used one at a time to change the 47 Sc activity values used in the fitting routine. For example, the branching ratio for t he 159.4 keV gamma ray was increased by one sigma fr o m 68.3% to 68.7%. This new branching ratio value was then used to find the activity of 47 Sc, and the fitting algorithm was run again with these new activity values. This process was then repeated by sett ing the branching ratio back to 68.3% and changing t h e efficiency by an amount equal to one sigma. Similarly, the times associated with each data point ( i.e. , the time from the separation of 47 Ca and 47 Sc) were varied up or down by one sigma and were used in the fit algorithm. The difference between the hal f - lives found with each of the parameters varied by one sigma were added in quadrature with the uncertainties in the half - life values for 47 Ca and 47 Sc found with the true values for the parameters . The uncertainty in each of the three 47 Ca gamma - ray bran c hing ratios was found in a similar way as described above for the 47 Ca and 47 Sc half - lives using the Bateman equation. The branching ratios were calculated with only the counting statistics for each gamma - ray peak as the uncertainty. For example, the coun t ing statistics for the 159.4 keV peak and the 489.2 keV peak were the only uncertainties propaga ted through the calculation to find the 489.2 keV branching ratios. The weighted average and associated uncertainty for each gamma - ray branching ratio was then found using the errors resulting from these counting statistic uncertainties. Additional uncerta inties that were common across all data points were used to 207 vary their associated constants by one sigma, as explained previously, to observe the effect on the branching ratio weighted average. These uncertainties in the efficiency of the detector at each energy, in the decay constants of 47 Ca and 47 Sc, the 159.4 keV branching ratio, and in the exact time of separation of 47 Ca and 47 Sc. The difference between ea c h of these new weighted averages and the original weighted average was assigned as the uncertain ty introduced by each parameter used to find the branching ratios. A total uncertainty was calculated by adding the square of all these uncertainties and the u n certainty in the original weighted average in quadrature. Each of the uncertainties and their so urces are given in the Results and Discussion section. 8 .3 Results and Discussion Since the measurement of the three 47 Ca branching ratios require the use of t he half - li ves of 47 Ca and 47 Sc, these half - lives were measured with LSC. The sample used for 47 S c was a purified fraction that contained levels of 47 Ca below the limit of detection using HPGe gamma - ray measurements, so the entire channel window for the LS C spectra was considered for the half - life measurement of 47 Sc. As 47 Ca decays continuously to 47 Sc, producing a purified 47 Ca fraction containing no other radionuclides over a long period of time was not possible. Therefore, the LSC channel range us ed for the 47 Ca half - life measurement was restricted to a region containing only signals from the decay of 47 Ca. A portion of the earliest LSC data points collected for three 47 Ca samples was not included in the determination of the half - life of 47 Ca due to ran d om coincidences events in the data with the highest activity level . The number of random coincidence events in an LSC spectrum for a pure 47 Ca source has been shown to be proportional to the square of the activity 208 Table 8 . 1 : Half - L ives of 47 Ca and 47 Sc The reduced chi - squar ed value ( i.e. , the chi - squared value divided by the degrees of freedom for each fit) is given to provide information on the goodness of fit for each half - life. Radionuclide Source Half - L i fe (days) Reduced Chi - S quared (DOF) 47 Ca Evaluated half - life [51] 4.536(3) - LSC, sample 1 4.53(1) 2.5 7 (23) LSC, sample 2 4.532(9) 1.90 (28) LSC, sample 3 4.53(1) 1.56 (3 6 ) Average of LSC measurements 4.53 1 (5) - HPGe, 489 keV gamma ray 4.54(2) 1.38 (39) HPGe, 807 ke V gamma ray 4.52(3) 1.35 (39) HPGe, 1297 keV gamma ray 4.53(1) 1.19 (39) Average HPGe direct measurement 4.534(3) - HPGe, 159 keV gamma ray 4.6(1) 1.07 (38) 47 Sc Evaluated half - life [51] 3.3492(6) - LSC 3.3 50 (2) 1.03 (39) HPGe - 159 keV gamma ray 3.3(1) 1.07 (38) HPGe - 159 keV gamma ray; 47 Ca decay constant fixed 3.349(8) 3.91 (39) of 47 Ca [85] . In the same study, it was also shown that the number of random coincidences between radiation from both 47 Sc and 47 Ca is proportional to the product of the activities of the two radionuclides. Since these effects are most prom inent a t higher activit y levels , the earliest data points for each of the samples were systematically removed and the resulting half - life for each sample was seen to approach the accepted 47 Ca half - life value as more data points were removed . The maximum n umber of data points that produced a half - life within one standard deviation of the ac cepted half - life value were used in the verification ( i.e. , data points 17 to 41, 11 to 40 and 3 to 40 were used for samples 1 to 3, respectively). These data points and the fitt ed decay curves for each sample are shown in Figure 8 .2a and the resulting hal f - lives are given in Table 8 . 1. 209 Table 8 . 2 : Error Budget for Half - lives Found from 159.4 keV Gamma - Ray Peak The uncertainty related to the separation time resulted from the ap proximately 8 minutes required to separate 47 Ca and 47 Sc on the DGA resin. The separation time was chosen as the midpoint of this separation length and the uncertainty was assigned as half of the separati on length. A dash in the table indicates that the so urce of uncertainty was not relevant for the measurement whereas a value of zero indicates that the source of uncertainty was considered for the measurement but resulted in no additional un certainty. In p articular, the efficiency and branching ratio for 15 9.4 keV were constants with uncertainties relevant for finding the half - life of 47 Sc. However, these uncertainties resulted in no change in the fitted half - life since they uniformly shift a ll data points and do not affect the decay rate. Source of Uncert ainty Half - L ife Uncertainty 47 Ca 47 Sc 47 Sc with Constant 47 Ca Half - L ife Counting Statistics and Baseline Correction 0.084 0.066 0.0067 Efficiency at 159.4 keV 0.040 0.024 0.0000 Branching Ratio - 159. 4 keV 0.063 0.044 0.0000 Separation Time 0.085 0.064 0.0032 De cay constant of 47 Ca - - 0.0026 Total 0.14 0.10 0.0078 Figure 8 . 2 : Half - L ives for 47 Ca: LSC Count Rates The LSC counting rates are shown fo r the data p oints used to fit the half - lives of 47 Ca (a) and 47 Sc (b). The fitted decay curves fou nd for each sample are overlaid. Error bars from the counting statistics are given for each point but are mostly smaller than the size of the data points. 210 T he 47 Sc samp le that was measured with LSC was not influenced by random coincidences to the same extent as the 47 Ca samples due to a lower activi ty and the presence of only one radionuclide with a relatively few emissions per decay . Therefore, all the LSC m easurement s collected for the 47 Sc sample were used to verify t he half - life of 47 Sc as shown in Figure 8 .2b. The half - life found by fitting these data points with Equation 8 . 2 is given in Table 8 . 1 and is only 0.03% larger than the accepted half - life for 4 7 Sc. The h al f - lives of 47 Ca and 47 Sc measured with LSC support the accuracy of the accepted 47 Ca and 47 Sc half - lives. Therefore, the evaluated values of 4.536(6) and 3.3492(6) days for 47 Ca and 47 Sc , respectively, were used to find the three 47 Ca branching ratios [51] . The half - lives with w hich the 489.2, 807.9, 1297.1, and 159.4 keV gamma - rays were found to decay in the gamma - ray spectra for the 47 Ca/ 47 Sc sample also verified the purity of these gamma - ray peaks since they were in agreement with the evaluated half - lives for these radionucl id es . Figure 8 .3a shows the integrated number of counts for the three 47 Ca gamma rays and a simple exponential fitted function for each peak. Since the background - corrected integrated counts were used, the uncertainties only included con tribution s from cou nt ing statistics and the background correction for each data point. The activity of 47 Sc resulting from th e integrated counts in the 159.4 keV peak across the gamma spectra are shown in Figure 8 .3b. The Bateman equation was used to fit the se data points fi rs t with the half - life of both radionuclides and the initial activity of 47 Ca as variables and then again with only the half - life of 47 Sc and the initial activity of 47 Ca as variables. Only uncertainties from the background c orrecti on and countin g statisti cs were used in both fitting routines. As described previously, 211 Figure 8 . 3 : Half - L ives for 47 Ca and 47 Sc: Gamma - Ray Spectroscop ic Peaks The integrated number of counts for the 489.2, 807.9, and 1297.1 keV gamma - ray lines for 47 Ca (a) and the 47 Sc acti vity found with the 159. 4 keV gamma ray (b) are shown as a function of time for each gamma spectrum. The 47 Ca data points were fitted with a simple exponential for the 47 Ca gamma rays and the 47 Sc da ta point were fi tted with a Bateman equation. Error bars were based on counting s tatistics and background corrections and are mostly smaller than the data points. the uncertainties in the parameters used to find the activity of 47 Sc for each spectrum wer e systematically varied to find the uncertainty contribution from each consta nt (Table 8 . 2). With the dead time of the detector at 1% or less for each spectrum in this analysis, a dead time correction was not necessary for accurate results. The four resul ting values fo r the 47 Ca half - li fe and two values for the 47 Sc half - li fe found in this analysis are given in Table 8 . 1. Each measurement agrees quite well with the corresponding evalua ted value with the error bar for the half - lives found using the Bateman equation being l arger at 3% for the half - life of 47 Sc and 2% for the half - life of 47 Ca when these values were both used as variables. When the half - life of 47 Ca was fixed at the accept ed value, however, the half - life of 47 Sc was found with only a 0.2% unce rtainty. These r esults confirm 212 that, as expected, the 489.2, 807.9, and 1297.1 keV gamma rays come from 47 Ca and the 159.4 keV gamma ray comes from 47 Sc with little to no interferences from other sources. With confirmation that both sets of half - lives fo r 47 Ca and 47 Sc agree with the evaluated half - lives, the gamma spectra of the 47 Ca/ 47 Sc sample were analyzed to extract the branching ratios of the three main 47 Ca gamma rays. The branching ratios for each gamma spectrum are shown for the 489.2 and 807.9 k eV gamma ray s in Figure 8 .4a and for the 1297.1 keV gamma ray in Figure 8 .4b. The average value across each set of data points is included a s a solid or dashed line through the points. The error bars of most points in each of the three sets overlap with th eir average v alu e demonstrating agreement among the values within each set. At later time points, a larger spread in th e values is observed due to lower count rates and therefore, larger statistical uncertainties. The error bars in Figure 8 .4 result from t he counting s tat istics from each gamma - ray peak as these are the only uncertainties that vary by data point under the p resent conditions . The other errors that influence the branching ratio value were Figure 8 . 4 : Branchin g Ratios for Three Most Intense 47 Ca Gamma Rays The branching ratios observed in ea ch gamma - ray spectrum are given for the 489.2 and 807.9 keV gamma ray (a) and for the 1297.1 keV gamma ray (b). The solid or dashed line through the d ata poi nts indicates th e average branching ratio for each gamma - ray transition. Note that the error bars f or almost all the data points overlap with the respective average. 213 Table 8 . 3 : Error Budget for Three 47 Ca Br anching Ratios T he absolute uncertainty for each branching ratio as a percent is given for different sources of uncertainty. The uncertainty related to the separation time resulted from the approximately 8 minutes required to separate 47 Ca and 47 Sc on the DGA resin. The s eparation time was chose n as the midpoint of th e separation length and the uncertainty was assigned to be half of the separation length. Additionally, the efficiency uncertainty was found by varying the uncertainty at 159.4 keV and one 47 Ca gamma - ray energ y at the same time since these uncertainties were related. Source of Uncertainty Energy of Gamma Ray (keV) 489.2 807.9 1297.1 Counting Statistics 2.06E - 2 2.78E - 2 1.13E - 1 Efficiency 1.84E - 2 1.83E - 2 5.85E - 2 Decay Constant - 47 Ca 1.97E - 3 1.92E - 3 2.18E - 2 Decay Constant - 47 Sc 5.6E - 4 5.7E - 4 5.4E - 3 Branching Ratio - 159.4 keV 4.2E - 2 4.2E - 2 4.4E - 1 Separation Time 7.0E - 3 7.1E - 3 2.6.7E - 2 Total 5. 1 E - 2 5.4 E - 2 4.6 E - 1 incorporated in the final error of the weighted average value. All uncertainti e s for these mea surements are given in Table 8 . 3 where they are listed by the ir source. It is also vital to note for this measurement that the emission of the 489.2 and 807.9 keV gamma rays in the cascade from the 1297.1 keV excited state is isotropic wit h angular co rrel ation coefficients of A 2 = - 0.050(6) and A 4 = 0 [84,86] . With this isotropic distribution, the probability of simultaneously detecting both the 489 and 807 keV emission from the same decay event ( i.e. , true coinc i d ence summi ng) is on the order of the detection efficiency for these gamma rays (i.e., 1E - 4 for 807.9 keV and 3E - 4 for 489.2 keV). Since this probability is small compared to the uncertainties in these measurements, no correction was made to the branching r atios for this coincidence probability. The final values for the branching ratios of the 489.2, 807.9 and 1297.1 keV gamma rays are 7.17(5)%, 7.11(5)%, and 75.0(5)%, respectively. These values differ significantly from the 214 current evaluated branching r at ios of the se g amma rays : 5.9(12)%, 5.9(12)%, and 67(13)% , respectively. Additionally, the precision of the present branching ratios with uncertainties of less than 1% for each is much higher than that of the evaluated values . With these remeas ured branch in g ratios f or 4 7 Ca, other minor gamma - ray branching ratios can be adjusted in the 47 Ca decay scheme ( see Figure 8 .5). In particular, the 41.1, 530.6, 767.1, and 1878 keV gamma rays have been measured previously through a ratio to the more intense gamma ra ys in the 47 Ca d ecay scheme. Additionally, the 731.6 and 1147 keV gamma rays have been inferred through intensity bal ancing from the 1878 keV excited state. The branching ratio as an emission percent for 47 Ca decays based on the 489.2, 807.9, and 1297.1 ke V branchin g ra tios based on this work are given in Table 8 .4 and Figure 8 .5. The beta decay from the ground state of 47 Ca to both the ground state and the 1297.1 keV excited state of 47 Sc are currently reported with high uncertainties. These values can b ot h be found wit h far lower uncertainty using the 489.2 and 1297.1 keV branching ratios reported here. The relationshi p between key nuclear data in the decay of 47 Ca is shown in the decay Table 8 . 4 : Minor Gam ma - Ray Branc hing Ratio Values Gamma - Ray Energy (keV) Measurement Description Ratio Evaluated Branching Ratio (%) Branching Ratio from This Work (%) 41.1 Relative to intensity 100 for 1297.1 keV gamma ray [87] 0.0085(10) 0.0056(13) 0.0064(8) 530.6 Ratio with 489.2 keV gamma ray [86] 0.0146(10) 0.086( 18) 0.105( 7) 731.6 Intensity balanced 1878 keV energy level [51] - 0.011(3) 0.009(2) 767.1 Ratio with 807.9 keV gamma ray [86] 0.0294(10) 0.18(4) 0.209(7) 1147 Intensity balanced 1878 keV energy level [51] - 0.011(3) 0.009(2) 1878 Relative to intensity 100 for 1297.1 keV gamma ray [87] 0.038(4) 0.025(6) 0.028(3) 215 scheme in Figure 8 .5. Since on ly the 489. 2, 5 30.6 and 1297.1 keV gamma rays are emitted from the 1297.1 keV excited state for 47 Sc, the beta decay branching ratio is equal to the sum of the branching ratios of these three gamma rays. Internal conversion do es occur to de - exc ite this ene rgy level, but the rates are negligible compared to even the small uncertainties in these three gamma - ray branching ratios [51] . With the branching ratios measured in this work and updated 530.6 keV branching ratio given in Table 8 .4, the beta decay b ranching ratio fro m the ground state of 4 7 Ca to the 1297.1 keV excited state of 47 Sc is found to be 82.2(5)% [86] . This updated beta decay branching ratio can be used with the previously reported branching ratios for beta decay from the gro und state of 47 Ca to the 766.8 and 1878.2 keV excited states of 47 Sc (i.e., 0.087(3) and 0.037(8)%, respectively) [ 51,86 88] . Int ensity balancing with these values gives a beta decay branching ratio of 17.7(5)% f or the ground state to ground state beta decay Figure 8 . 5 : Decay Scheme of 47 Ca Beta Decay to 47 Sc The four beta decay bra nches for 47 Ca and the subsequent gamma - ray emissions from the excited energy level s in 47 Sc are given. The 489.2, 530.6, and 1297.1 keV gamma rays are the dominant emissions from the 1297.1 keV excited state in 47 Sc. The branching ratios fo r each beta and gamma - ray emission branches are indicated in the figure as a percent. 216 of 47 Ca to 47 Sc. The branching ratios for the beta decay of 47 Ca to the 1297.1 keV excited state and the ground state of 47 Sc given here fa ll within the large uncertaint y of the ac cept ed values but are much more precise. The values for all four beta decay branches are given in the decay scheme in Figure 8 .5. 8 .4 Conclusion Using a sample of pure 47 Ca produced at the NSCL, th e half - life of 47 Ca and its daughter, 47 Sc, we re measured with LSC and gamma - ray spectroscopy and were found to agree with the current evaluated half - li fe values . Using the ingrowth of 47 Sc and well - known nuclear data for this radionuclide, the branching r atio for the 489. 2, 807.9, and 1297.1 keV 47 Ca gamma rays were remeasured as 7.17(5)%, 7.11(5)%, and 75.0(5)%, respectively. These values are more precise than the current accepted branching ratios for these gamma rays. Additionally, the beta decay branchi ng ratio from the ground state of 47 Ca to the 1297.1 excit ed s tate and the ground state of 47 Sc were found to be 82.2(5)% and 17.7(5)%, respectively. Although these values fall within the uncertainty range for the currently reported values, they have a mar ked increase in p recision. These updated value s will allow for more accurate and precise quantification of 47 Ca when using 47 Ca to generate 47 Sc for nuclear medicine applications. 217 Chapter 9 : General Discussion 9 .1 Introduction T he results and techniques from the three 48 Ca irradiations presented here a re be critic ally compared and evaluated in this chapter to provide overall conclusions from the three experiments and identify some of the best techniques for future isotope harvest ing experiments. The production rate measureme nts, separation techniques, a nd stable el emen tal analysis results will be examined in this chapter to provide clarity for the recommended next steps for these aspects of the project. 9 .2 Production Rate of 47 Ca The production rate of 47 Ca in the flowing - water isotope harvesting target was m easured duri ng t wo different experiments as 2.0(4)% in Chapter 4 Section 4.2.3 and 1.6 7 ( 4 )% in Chapter 6 Section 6.3.1 , and these two production r ates are in agreement . In the first case, the total activity of 47 Ca coll ected on the cation exchange resin be d and water samp les or remaining in the water after collection were corrected back to the end of the irradiation and compared to the recorded beam current measurements to find a production rate. For the second measureme nt, the activity of 47 Ca collected in water sampl es a t three different points during the irradiation was extrapolated to find the total activity in the water system and then compared to the recorded beam current m easurements. Since the measured production rates agree despite the different met hods used to fin d the values, additional confidence is provided in the accuracy of the production rate measurement. 218 Additionally, the activity and production rate of 47 Ca fou nd in the first irradiation with a 48 Ca bea m can be recalculated using the remea sured branch ing ratios for the three main gamma rays following the decay of 47 Ca presented in Chapter 8. This results in an updated production rate of 1.71(5) % for this measure ment. The average production rate based on these two independent measurements ( i .e. , 1.71(5) % fr om Chapter 4 and 1.67(4)% from Chapter 6) is 1.69(2)% . A calculated 47 Ca activity for each irradiation in this study can be found with th e average production rate, the recorded beam currents throughout e ach irradiation , and Equation 4 .1 . Th ese calculat ed a ctivities are given in Table 9 .1 below and are compared to the total measured 47 Ca activity for each irradiation. To provide an understand ing of the scale of the experiment, the integrated beam current i s also given. Since the half - life of 47 Ca is long com pared to the length of the irradiations, the ratio between the integrated beam current and the produced activity at the end of each irradi ation should be relatively consistent between experiments. The v alues shown in Table 9 .1 demonstrate some variabi lity in the comparison between the measured and calculated activity. These differences may be rooted in sources of Table 9 . 1 : Comparison of Measured to Calcula ted 47 Ca Activities Experiment Integrated Beam Curr ent (pnA h) Measured Activity (mCi) Calculated Activ ity (mCi) Percent Difference (%) Chapter 4 4.7 0. 086 (2) 0.08 24 ( 8 ) - 4. 0(2) Chapter 6 79.4 1.79(6) 1.4 4(1) - 19.5(8) Chapter 7 703 9.9(2) 12.08(2) + 22. 2(6) 219 uncertainty in the measurement of the beam cur rent that was impinged on the isotope harvesting target or the activity of 47 Ca produce d in eac h irradiation . For example, quantification of the beam current including the calibration of the non - intercep ting probes in the beam line and calib ration of t he t arget current readings based on the non - intercepting probe readings may have larger unc ertainti es than estimated . Some of the difference between the measured and calculated activity might be due to the a ccuracy with which high activity measu rements wer e ma de with gamma - ray spectroscopy. The non - standard geometries used for many measurements, such as samples of tens to hundreds of milliliters of water or samples collected on columns, may have introduced in accuracies to the measurements. In par ticular, me asur ements of the activities collected on cation exchange resin beds were difficult to meas ure. The se activities were quite high, and aliquots of the activity could not be used for quantification. As a resul t, the collection resin beds were measu red at dis tanc es far from the detector face and even so had dead times of 5 to 10%. Additionally, vari ations i n each experiment may also contribute to the disparity between measured and calculated 47 Ca activities in these irradiations. For instance, positi oning of the beam spot on the flowing - water target may change the production rate in the target. The e nergy pr ofile of the beam changes if the beam is lower on the target and passes through the internal wall in the t arget or if the beam spot is higher and does not pass through the internal wall. While the difference between the measured and predicted 47 Ca activiti es was up to 22% of the measured activity across these three experiments , the measured production rate pro vided adequate accuracy for predicting t he scale of p roduced activity in future experiments. This should help with both experimental and safet y planni ng in future, higher intensity experiments. 220 9 .3 Separation Procedures for the Purification of 47 Ca In Ch apter 4 , three separation schemes were explored with all three produc ing high puri ty 47 Ca in 3 to 5 M HCl. While the separation method involving AG MP - 50 and an HCl/methanol gradi ent was eliminated due to some impracticalities of this method ( i.e. , high vo lumes and the use of methanol), the two other methods , 1) AG MP - 50/HC l and 2) DGA HNO 3 /HCl , were shown to be viable options. The first of these methods was identified as the best choice due to the reported instability of the DGA resin when exposed to high levels of radiation and with repeated use. After performing several AG MP - 50/ HCl separations with the harvested 47 Ca (see C hapter 6 Section 6 .2.5 and 6 .3.3 and Chapter 7 Section 7 .2.4 and 7 .3.2), the suitability of th e method became more apparent . The elu tion of 47 Ca from the cation exchange collection resin bed was optimi zed with 1 an d 2 M HCl to carry out the separation with out an evaporation step . Following elution of about 85% of the collected 47 Ca, the solution was diluted to a lower acid level that w ould be suitable load solution s for the gradient AG MP - 50/HCl separat ion. A re lati vely large volume load solution resulted f rom these optimizations since about 25 mL of 2 M HCl wa s required to remove about 85% of the collected 47 Ca and a 0.2 M HCl load sol ution provided the highest yield of 47 Ca with 99.9% purity. Together, this res ulte d in a 250 mL 0.2 M HCl load phase, which is a high volume compared to the DGA HNO 3 /HCl method. F urther optimization including lowering the flow rate and increasing the lengt h of the column could allow for a higher acidity and therefore a lowe r volume load phase. The elution of the cation exchan ge collection resin bed in Chapter 4 (Section 4 .2.4 and 4 .3.2) demonstrated that a rinse step of about 70 mL of HNO 3 would remove al most all the 47 Ca 221 collected on the resin bed. This volume can be load ed direct ly o nto a DGA resin for purification of 47 Ca i n the DGA HNO 3 /HCl separation and is much smaller than the volume used for the AG MP - 50/HCl separations in later experiments. One ad ditional issue that could arise from using the DGA HNO 3 /HCl method in future e xper iments is the low capacity of the resin gi ven the significant amount of stable metal ion impuriti es observed in the two higher intensity 48 Ca experiments. Since a small shift was observed in the elution profile for the two AG MP - 50/HCl separat ions repo rted in Chapter 7 Section 7 .3.2 , this may a ffect the retention of 47 Sc on the DGA resin in the 4 M HC l load phase if the stable ion concentration is high enough to compete with 4 7 Sc for interaction with the resin . Another separation method was t ested bri efly in the second 48 Ca irradiation as disc ussed in Chapter 6 . The fourth fraction from the cation ex change collection resin bed was not purified with the AG MP - 50/HCl method. In stead, this portion was purified with the generator procedure and dem onstrated tha t the purification of scandium isotopes despite the presence of other radionuclides such as 7 Be, 42 Na, and 42,43 K in the generator solution (see Figure 6 .10). This result sho w ed that the first step in the generator procedure (DGA 1) could be u sed with 3 to 5 M HCl eluate from the cation exchang e collection resin bed to remove all scandium isotopes fro m the solution. A generator solution would be produced containing 7 Be, 42 Na, 42,43 K, and 45,47 Ca produced in the isotope harvesting system and 47 S c generat ed i n this solution. This proposed separati on would eliminate the need for a dilution step as require d in the AG MP - 50 separation and would be simpler than the DGA 3 M HNO 3 /HCl s eparation. The stable ions identified in Chapter 6 Section 6 .3.6 and Chapter 7 Sec tion 7 .3.5 would either harmlessly foll ow Ca 2+ through all the separation steps ( e.g. , Al, Co, Cu , Mg, Mn, Ni) or could be 222 removed from the 47 Sc product with the 3 M HNO 3 ste p added to the generator procedure ( e.g. , Fe and Zn). This separation could be car ried out by placing the collection colu mn and a small DGA column (50 - 70 mg) in parallel and using tens of mL of 3 M HCl rinsed through both columns to rapidly produce a suita ble 47 Ca/ 47 Sc generator solution. The drawback s to th e proposed sim plified s epar ation procedure are the presence of radionuclides other than 47 Ca and 47 Sc in the generator solut ion and the use of DGA resins in both this separation step and the generator procedure. The presence of additional radionuclides in the soluti on m eans that hig her activity levels would be present. This could lead to more extensive breakdown of the DGA resi n than with just 47 Ca and 47 Sc present in the generator solution. Potential c onsequences from this breakdown are reduced yield and purity resu ltin g from a sepa ration and the need to use a prefilter resin to remove traces of organic residue removed from the DGA resin. Additionally, such a contaminated generator solution might not pa ss requirements for the radionuclidic purify of a 47 Ca/ 47 Sc gener ator . While t his is not a scientific drawback for the generation of 47 Sc, it could limit the use of this method in further applications. As mentioned previously, this method would use DGA res in which is also used in the generator procedure. This repetition may reduce t he s eparation capacity of the overall method compared to using to different resins ( e.g. , using AG MP - 50 in the purification of 47 Ca and the DGA resin in the purification of 47 Sc ). Any of the three methods discussed here could be used to pro duce a high q uali ty 47 Ca/ 47 Sc generator. As experiments continue moving towards more intense conditions expected f or isotope harvesting during the regular operation of FRIB, it ma y become cle ar which of these three methods is the most advantageous. 223 9 .4 Sta ble Elemental Ana lysis The stable elements identified and measured in the two higher intensity 48 Ca irradiations (Chapter 6 Section 6 .3.6 and Chapter 7 Section 7 .3.5) showed that a significa nt amount of stable impurities were collected on the cation excha nge resin bed s co mpared to the amount of radionuclidic products of interest. These stable impurities will likely i ncrease as the intensity and duration of irradiations and the amount of expos ed metallic surface s in the isotope harvesting water system incre ase. The opti miza tion and selection of purification schemes should include measures to remove these stable ions as well as radionuclidic impurities. It is possible that the mass of stable ion s in the system could be mitigated by increased effort to recombi ne r adiolysis pro ducts in the system. By decreasing the buildup of hydrogen peroxide in the water, lower levels of stable ions may enter the water due to corrosion of metallic components. O ne additional interesting finding from the stable elemental analy sis of the ir radi at ed water is the presence of Al. From the level of other stable metal ions observed, one likely ld account for the high level of Al and the Mn ions. The other li kely source i n th e water system would be Al from the Ti64 alloy target shell which contains 6% Al and 4% V. Analys is of the stable ions collected on the anion exchange resin was not performed but would add support to either of these metals as the source of alu minum. An y va nadium ions that could have be removed from the target titanium alloy shell would have been colle cted on the anion exchange column. While the presence of vanadium on the anio n exchange column would support target material 224 degradation, its abse nce would ind icate 3003 aluminum alloy as the source. F uture irradiation experiments at high beam intensities could explore th e source of Al. 9 .5 Conclusion The series of experiments ca rried out in this work has explored many techniques for harves t ing 47 Ca from a hea vy - ion beam stop to provide a preclinical supply of 47 Sc in the future at FRIB. Such harvesting h as been shown to be possible and can provide a significant supply of 47 Sc , wh ich is currently diffi cult to produce for potential theranostic appli cations s uch as the targeted treatment of metastatic cancers. The techniques optimized in these experiments ha ve demonstrated the feasibility of this production route using simple procedu res with reproducible results. In addition, t he branching ratio s of t he gamma rays following the decay of 47 Ca were remeasured, supporting the use of isotope harvested radionuclid es for a wide range of applica tions . As harvesting efforts progress towards h arvesting at FRIB, the se techniques can be t ranslated to the conditio ns antici pate d with higher intensity irradiations and a somewhat different isotope harvesting water system at FRIB with the ultimate goal of providing large quantities of 47 Sc for further preclinical research. 225 APPENDICES 226 APPENDIX A: N UCLE AR DATA T able A 1 : Nuclear Data Used for Identification, Quantifi cation, and Localization of Radionuclides Produced with a 40 Ca Beam in a Flowing - W ater Target . [14,15,28,50 56] Nuclide Gamma - Ray Energy (keV) Branching Ratio (%) 7 Be 477.71 10.44(4) 18 F 511 193.46(8) 24 Na 1368.6 99.99 36(15) 28 Mg 4 00.6 8 36.6(10) 34m Cl 146.36 38.3(5) 43 Sc 372.72 22.5(7) 44 Sc 1157.020 99.9(4) 44m Sc 271.34 8 6.7(3) 47 Sc 159.47 68.3(4) 48 Cr 112.58, 308.4 96.0(20), 100(2) 48 V 983.47,1312 99.98(4), 98.2(3) 52 Mn 744.22, 935.59, 1434.29 90.0(12), 94.5( 13), 100.0(14) 227 Table A 2 : Nuclear Data Used to Quantify Radionuclides P ro duced By a 47 Ca Beam in a Flowing - Water Target 6,7,11,21,22,30 - 34 For the characteristic gamma - ra ys of 28 Mg, no uncertainty for th e branching ratio was reported so an uncertainty of 10% was assumed . Radionuclide Half - life Gamma - Ray Energy (keV) Branching Ratio (%) 24 Na 14.99 7 h 1368.6 99.99 36(1 5) 27 Mg 9.458 m 843.8 70.94(9) 28 Mg 20.915 h 400.6 36(4 ) 941.7 36(4) 1342.2 54( 5) 42 K 12.355 h 1524.6 18.08(9) 43 K 22.3 h 372.8 86.8(2) 396.9 11.85(8) 593.4 11.26(8) 617.5 79.2(6) 44 K 22.13 m 368.2 2.3(4) 651.4 3.0(5) 726.5 3.8 (6) 1024.7 7(1) 1126.1 8(1) 1157.0 58(9) 1499 .5 8(1) 1752.6 4.1(6) 45 K 17. 18 m 174.28 74(5) 1705.6 53(3) 47 Ca 4.536(3) d 489.2 5.9(12) 807.9 5.9(12) 1297.1 67(13) 44m Sc 58.61 h 271.2 86.7(3) 47 Sc 3.3492 d 159 .4 68.3(4) 48 Sc 43. 67 h 983.4 100.1(5) 1037.5 97.6(7) 1312.1 100.1(7) 228 Table A 3 : Nuclear Data and Geometry Correction Factor for Production Rate Measurement Radionuclide Half - life Gamma - Ray Energy (keV) Branchin g Ratio (%) Geome try Correction Factor 25 cm 50 cm 24 Na 14.997 h 1368.6 99.9936(15) 1.09(4) 1.13(5) 27 M g 9 .458 m 843.8 70.94(9) 1.14(3) 1.16(4) 1014.5 28.20(2) 1.13(3) 1.15(4) 28 Mg 20.915 h 400.6 36(4) 1.18(3) 1.18(3) 941.7 36(4) 1.13(3) 1 .15(4) 1342.2 54(5 ) 1.10(4) 1.13(5) 34m Cl 31.99 m 146.5 38.3(5) 1.20(3) 1 .19(3) 38 Cl 37.230 m 1643 32.9(5 ) 1 .08(4) 1.12(5) 39 Cl 56.2 m 250.3 46.1(1) 1.19(3) 1.19(3) 1091 2.42(8) 1.12(4) 1.14(4) 1267 53.6(13) 1.11(4) 1.13(5) 1517 39.3(10) 1.0 9(4) 1.12(5) 38 S 170 .3 m 1943 86.5(22) 1.05(5) 1.10(6) 42 K 12.355 h 1524.6 18.08(9) 1.09(5) 1.12(5) 43 K 22. 3 h 372.8 86.8(2) 1.18(3) 1.18(3) 396.9 11.85(8) 1.18(3) 1.18(3) 593.4 11.26(8) 1.16(3) 1.17(3) 617.5 79.2(6) 1.16(3) 1.17(3) 44 K 22.13 m 368.2 2.3(4) 1. 18(3 ) 1.18(3) 651.4 3.0(5) 1.16(3) 1.17(3) 726.5 3.8(6 ) 1.15(3) 1.16(4) 1024.7 7(1) 1.1 3(3) 1.15(4) 1126.1 8(1) 1.12(4) 1.14(4) 1157.0 58(9) 1.11(4) 1.14(4) 1499.5 8(1) 1.09(4) 1.12(5) 1752.6 4.1(6) 1.07(5) 1.11(6) 45 K 17.18 m 174.28 74(5 ) 1.19(3) 1.19(3) 957.5 7.6(5) 1.13(3) 1.15(4) 126 0.3 8.5(5) 1.11(4) 1.13(5) 143 4.5 4.2(3) 1.09(4) 1.13(5) 1705.6 53(3) 1.07(5) 1.11(6) 47 Ca 4.536(3) d 489.2 5.98(2) 1.17(3) 1.17(3) 807.9 6.0(3) 1.14(3) 1.16(4) 1297. 1 60.7(2) 1.10(4) 1.1 3(5) 44m Sc 58.61 h 271.2 86.7(3) 1.19(3) 1.18(3) 46 Sc 83.79 d 889.3 99.984(1) 1.14(3) 1 .15 (4) 1120.6 99.987(1) 1.12(4) 1.14(4) 47 Sc 3.3492 d 159.4 68.3(4) 1.19(3) 1.19(3) 48 Sc 43.67 h 983.4 100.1(5) 1.13(3) 1.15(4) 1037.5 97. 6(7) 1.12(3) 1.15 (4) 229 APPENDIX B: PREDICTED NUCLEAR REACTION PROBABILITIES Chapter 2: 40 Ca Irradiation For the fusion evaporation products found in the system, 43 Sc, 44,44m Sc, 48 V, 48 Cr, and 52 Mn, the reaction cross sections were predicted using PACE4 and LiFus in the p rogr am LISE++. In the Monte - Carlo simulation used in PACE4, each reaction was sampled 10,000 tim es ( i.e., the number 40 Ca and the target was set as 16 O. The optical model potential s were left as the def t h e cross section was simulated in PAC E4 over a range of lab energies. The spaces between the energy points were 50 MeV for cross - sections between 1 and 10 mb and 25 MeV for cross sections larger th an 1 0 mb. For any jumps in the cross section of greater th a n 50 mb, the spacing was decreas ed t o 12.5 MeV between energy points. Cross - sections less than 1 mb were not included in the total production calculations. Tables B1 - 5 give the cross - section valu es b y energy used for the production estimates. It should b e noted that these values were ge nera ted using a Monte Carlo simulation, meaning identical values may not be produced every time the simulation is run with the same settings. Lis Fus was used in LI SE++ - s m. With this mechanism selected, the projectile was set to 40 Ca, the residual was set to one of the radionuclides produced, and the target was set to 16 O. The excitation function for each reaction was output with steps of 11.72 MeV in beam lab energy (128 points between 0 and 1500 MeV). As w ith the PACE4 calculations, only cross - section values of greater than 1 mb were used to 230 determine production rates. Tables B6 - 11 provide the c ross - section value s fo r a specific energy that were used in the estimate. FR p roduction rates were simulated w ith LISE++ using a primary beam of 40 Ca 20+ at an energy of 137.59 MeV/nucleon and an intensity of either 0.11 pnA or 0.42 pnA for the low and high intensity beam se ttin gs, respectively. The target was set to 0.7 mm of Ti6Al 4 V followed by an 11 mm water str ippe r layer, and finally, a thick material block of iron. This set - up allowed for a slight attenuation of the beam intensity and energy in the Ti alloy target shell bef ore the beam entered the water. The water layer thickne s s was optimized by considering t hat the water must be thick enough to allow all the possible nuclear reactions to occur in the water while thin enough to allow the particles to p ass through the wa ter layer into the iron material, which serves as a particl e detector in the simulation. Bec ause the fragmentation reactions occur at high energies, this restriction did not strongly influence the calculation. In practice, the optimum wat er thickness was f ound by using a range of water thicknesses and finding the t hickness that gave the highest d etec tion of the fragments of interest in the material layer. 231 Table B 1 : Cross S ection D at a from PACE4 for the P roduction o f 43 Sc through the 16 O + 40 Ca F usion E vapora tion R eaction. Energy of 40 Ca Be am (MeV) Cross Sect ion (mb) 200 2.27 250 10.6 300 4.54 325 8.16 350 17.9 375 33.3 400 41.3 425 49.9 450 43.5 475 33.3 500 27.4 525 27. 3 550 36.2 575 38.6 600 46.8 625 43.8 650 41.2 675 34.4 700 27.4 725 25.1 750 18.6 775 18.6 800 14. 7 825 14.7 850 12 .6 875 9.85 900 9.05 950 5.66 1000 2.38 232 Table B 2 : Cross S ection D ata from PACE4 for the P rodu ction of 44 Sc through the 16 O + 4 0 Ca F usion E vaporation R eaction Energy of 4 0 Ca Beam (MeV) Cross Section (mb) 275 7.15 300 27. 3 325 52.9 350 65.7 375 68 400 51.1 425 35.2 450 32.9 475 45.4 500 65.5 525 81.7 550 82.5 575 75.8 600 59.1 625 50 .7 650 41 675 37.3 700 30.9 7 25 3 2.8 750 26 775 27.4 800 24.9 825 18 850 14.4 875 12 900 6.34 950 2.38 233 Tabl e B 3 : Cross S ection D ata from PACE4 for the P roduction of 48 V through the 16 O + 40 Ca F usion E vaporati on R eaction Energy of 40 Ca Beam (MeV ) Cross Section (mb) 200 13.7 212.5 41 .1 225 85.4 237.5 141 250 193 275 229 300 188 312.5 150 325 116 350 69.9 375 56.1 400 73.3 425 86.7 450 94.1 475 91.4 500 74 525 49.7 550 32.2 575 15.9 600 10.4 625 4.76 650 2.72 234 Table B 4 : Cross S ection D ata from PACE4 for the P roduction of 48 Cr through th e 16 O + 40 Ca F usion E vaporation R eaction Energy of 40 Ca Beam (MeV) Cross Section (mb) 100 1.21 150 9.97 200 2.04 225 7.72 250 29.6 275 51.7 300 55.1 32 5 38 .3 350 22.7 375 14.4 400 12.1 425 16 .3 450 15.7 475 19.6 500 16.1 525 12.5 550 7.82 575 6.46 600 2.04 625 1.59 235 Table B 5 : Cross S ection D ata from PACE4 for the P rod uction of 52 Mn through the 16 O + 40 Ca F usion E vaporation R eaction Energy of 40 Ca Beam (MeV) Cross Section (mb ) 125 6.11 137.5 43.9 150 135 162.5 243 175 317 200 314 212.5 246 225 194 237.5 123 250 83.2 262.5 52.9 275 29.4 300 8.62 236 Table B 6 : Cross S ection D ata from Li sFus for the P roduction of 43 Sc t hrough the 16 O + 40 Ca F usion E vaporation R eaction Energy of 40 Ca Beam (MeV) Cross Section (mb) Energy of 40 Ca Beam (MeV), continued Cross Section (mb), continued Energy of 40 Ca B eam (MeV), continued Cross Section (mb), con tinued 316.43 1.26 644.57 39.11 972.72 73.33 328.1 5 2.37 656.29 38.82 984.44 68.92 339.87 7.64 668.01 35.53 996.16 64.52 351.59 13.28 679.73 31.39 1007.88 60.13 363.31 18.59 691.45 29.68 1019.60 55.39 375.0 3 23 .81 703.17 29.53 1031.32 50.03 386.74 2 4.77 714.89 30.34 1043.04 43.99 398.46 22.93 726.61 32.37 1054.76 34.95 410.18 20.24 738.33 34.26 1066.48 25.97 421.90 16.48 750.05 35.74 1078.20 24.34 433.62 13.11 761.77 37. 48 1089.92 22.72 445.34 10.72 77 3.49 41.69 1101.64 19.49 457.06 8.73 785.21 45.90 1113.36 15.93 468.78 9.69 796.93 47.93 1125. 08 13.18 480.50 10.66 808.65 49.77 1136.79 10.84 492.22 13.57 820.37 53.77 1148.51 8.90 503.94 16.58 832.09 58.48 1160.23 7. 40 515.66 21.17 843.81 65.84 117 1.95 6.05 527.38 26.19 855.53 75.13 1183.67 5.05 539.10 31.11 867.25 81.67 1195.39 4.09 550.8 2 35.96 878.96 84.12 1207.11 3.42 562.54 38.30 890.68 85.91 1218.83 2.75 574.26 37.63 902.40 85.43 1230.55 2.29 585.98 37.77 914.12 84.92 1242.27 1.84 597.7 0 39 .89 925.84 83.68 1253.99 1.51 237 Table B 6 (cont d) 609.42 41 .63 937.56 82.43 1265.71 1.23 62 1.14 40.53 949.28 79.92 1277.43 1.01 632.85 39.44 961.00 77.23 238 Table B 7 : C ross S ection D at a fr om LisFus for the P roduction of 44 Sc through the 16 O + 40 Ca F usion E vaporation R eaction E ne rgy of 40 Ca Beam (MeV) Cross Section (mb) Energy of 40 Ca Beam (MeV), continued Cross Section (mb), continued Energy of 40 Ca Beam (MeV), continue d Cross Section (mb) , continued 292.99 1.68 574.26 17.02 855.53 38.85 304.7 1 3.59 585.98 16.13 867.25 35.96 3 16.43 5.30 597.70 15.24 878.96 32.74 328.15 6.11 609.42 12.74 890.68 29.36 339.87 6.66 621.14 10.23 902.40 25.91 351.59 5.97 632.85 11.01 914 .12 23.39 363.3 1 5. 20 644.57 11.87 925.84 21.27 375.03 4.00 656.29 13.02 93 7.56 19.15 386.74 2.88 668.01 14 .1 8 949.28 17.05 398.46 2.15 679.73 16.47 961.00 15.08 410.18 1.60 691.45 18.85 972.72 13.19 421.90 2.15 703.17 21.32 984.44 11.53 433.62 2.95 714.89 23.80 99 6.16 10.03 445.34 5.53 726.61 24.32 1007.88 8.66 457.06 8.2 5 738.33 24.58 1019.60 7.39 468. 78 12.18 750.05 27.47 1031.32 6.25 480.50 15.92 761.77 30.78 1043.04 5.23 492.22 17.63 773.49 33.86 1054.76 4.36 503.94 19.29 785.21 36.90 1066 .48 3.64 515.66 20. 21 796.93 36.84 1078.20 2.90 527.38 21.07 808.65 36.07 1 089.92 2.13 539.10 20.87 820.37 38 .41 1101.64 1.60 550.82 20.61 832.09 41.57 1113.36 1.31 562.54 18.81 843.81 40.81 1125.08 1.03 239 Table B 8 : Cro ss S ection D ata from LisFus for the P roduction of 48 V through the 16 O + 40 Ca F usion E vaporation R eaction Ener gy of 40 Ca Beam (MeV) Cross Section (mb) Energy of 40 Ca Beam (MeV), continued Cross Section (mb), continued Energy of 40 Ca Beam (MeV), continued C ross Section (mb ), c ontinued 210.95 4.96 363.31 9.20 515.66 25.39 222.67 10 .94 375.03 16.74 527.38 21.11 23 4. 39 24.75 386.74 24.44 539.10 17.41 246.11 36.77 398.46 33.09 550.82 13.75 257.83 41.37 410.18 41.51 562.54 11.14 269.55 43.27 421.90 48.52 57 4.26 8.55 281.2 7 33 .88 433.62 54.42 585.98 6.83 292.99 24.76 445.34 52.43 5 97.70 5.11 304.71 16.82 457.06 5 0. 30 609.42 3.93 316.43 9.79 468.78 47.10 621.14 2.75 328.15 6.88 480.50 43.56 632.85 2.05 339.87 4.77 492.22 36.54 644.57 1.38 351.59 6.52 50 3.94 29.71 656.2 9 1. 01 240 Table B 9 : Cross S ection D ata from LisF us for the P roduction of 48 Cr through the 16 O + 40 Ca F usion E vaporation R eaction Energy of 40 Ca Beam (MeV) Cross Section (mb) Energy of 40 Ca Beam (MeV), continue d Cr oss Section (mb), continued Energy of 40 Ca Beam (MeV), co ntinued Cross Section (mb), conti nu ed 105.48 1.66 304.71 46.58 503.94 81.33 117.20 2.03 316.43 27.61 515.66 67.60 128.91 2.24 328.15 22.79 527.38 54.00 140.63 1.35 339.87 20.6 4 539.10 42.54 152. 35 0.57 351.59 31.26 550.82 31.31 164.07 0.32 363.31 44. 55 562.54 26.14 175.79 0.45 375. 03 71.56 574.26 20.99 187.51 2.41 386.74 97.72 585.98 16.82 199.23 9.68 398.46 119.12 597.70 12.65 210.95 40.11 410.18 138.62 609.42 9.68 222. 67 71.99 421.90 146. 12 621.14 6.71 234.39 109.94 433.62 152.34 632.85 4.84 246.11 139.76 445.34 149.30 644.5 7 2.99 257.83 135.91 457.06 144.95 656.29 2.25 269.55 127.02 468.78 129.61 668.01 1.57 281.27 97.12 480.50 114.17 679.73 1.17 292.99 68.69 492 .22 97.65 241 Table B 10 : Cross S ect ion D ata from LisFus for the P rod uc tion of 52 Mn through the 16 O + 40 Ca F usion E vaporation R eaction Chapter 3 : 48 Ca Irradiation The program LISE++ was used to verify the relative product ion rate of 47 Ca through fragmentation reactions with a 140 a nd a 189 MeV/nucleon 48 Ca. To fin d these values, the following settings where used in the program with both beam energies . The components in the beam line were (in order) a 0.57 m m thick target o f Ti 64 alloy (86% Ti, 10% Al, and 4% V by stoichiometry, 4.43 g/cm 3 ), a stripper layer of wate r (67% H, 33% O by stoichiometry, 1 g/cm 3 ), and a material layer of iron. This set up represents the iso tope harvesting target as a thin shell of Ti64 alloy and a thi ck internal layer of water. As the beam travels through t he water layer, 47 Ca is formed th ro ugh fragmentation reactions on 1 H and 16 O/ 18 O nuclei. The product continues with a forward momentum an Energy of 40 Ca Beam (MeV) Cross Section (mb) 128.91 1.99 140.63 13.01 152.35 32.64 164.07 98 .41 175.79 152.92 187.51 154.53 199.23 146.06 210.95 93.7 0 222.67 47.54 234.39 27.30 24 6. 11 10.15 257.83 5.76 269.55 2.05 281.27 1.16 242 the pr ogra m. Therefore, the thickness of the water layer was optimi zed for each beam energy, as it n ee ds to be thick enough to allow for the beam to complete all possible fragmentation reactions and thin enough to allow for the products to have e nough energy to esca pe and implant in the material layer. The water layer was optimized to 14.5 mm and 25.6 mm f or the 140 and 189 MeV/nucleon beam, respectively. These settings and a beam intensit y of 4. 992 X 10 11 pps for both beam energies for were used to find the perc ent primary beam conversion to 47 Ca. 243 APPENDIX C: STABLE ELEMENTAL ANALYSIS Ta bl e C 1 : Wavelengths Used for Identification and Quantification of Stable Elements in System Water Element Wavelength An alyzed (nm) Al 396. 152 Fe 238.204 Ni 231.604 Si 212.412, 251.611 P 213.6 18 Ca 317.933, 396.847, 422.673 K 766.491, 769.897 Mg 279.553 Na 568.821, 589.592 B 249.772 Zn 213.857 S 181.972 244 Table C 2 : ICP - OES Instrum ent Settings Used in Chapter 5 and 6 The boost and the snout purg e were both turned on during al l calibrations and sample runs. Also, the Instrument Setting Value Set Pump speed 11 rpm Pump upta ke rate (mL/min) 2 5 mL /min Pump inject rate (mL/min) 7.0 mL/min Valve uptak e delay 6.1 seconds RF power 1 .3 kW Bubble injection time 0 seconds Preemptive rinse time 0 seconds Uptake delay 0 seconds Rinse time 0 seconds Read time 8 seconds Viewing mode SVDV Nebuliz er f low 0.70 L/min Plasma flow 12.0 L/min Aux flow 1.00 L /min 245 APPE N DIX D: PRODUCTION RATE MEAUSREMENTS IN CHAPTER 5 Water Volume Measurements The volume in the system was carefully measured to accurately find the total activity in the system when each water sample was taken. Throughout the irradiation, every water sam p le taken was carefully weighed and the mass and time of removal were recorded. Following the experiment, the total remaining water volume in the system was measured t wice in the same manner. Together, this information gave the total water volume in the sy s tem at the time in which each water sample was withdrawn for the production rate measurement. This is important because the activities for each r adionuclide measured in t he 250 mL water samples were multiplied by a factor to account for the total activity in the isotope harvesting system water. The water volume remaining in the system after the irradiation was measured by quantifying the dilution factor of a CaCl 2 sam ple added to the system. 100 mL of a 7500 ppm Ca 2+ solut ion was spiked into the water res e rvoir of the isotope harvesting system. The water was circulated through all components of the system to thoroughly mix the spiked Ca 2+ . Mixing w as verified by steady rea dings on the conductivity probes used througho ut the system to indicate the level of d issolved ions in the water. A water sample of about 40 mL was then removed and weighted. This dilution method was repeated after using the large mixed bed resins to r educ e the conductivity of the system water to a ba ckgrou nd of about 200 nS/cm. Another sa m ple was taken to measure the baseline concentration of Ca 2+ before the second measurement. 246 These diluted samples from the system as well as a sam ple of the original C aCl 2 solution were diluted with 1.4% HNO 3 and the Ca 2+ c oncentration was quantified with I CP - OES. Standards of 0 - 10 ppm Ca 2+ were produced in 1.4% HNO 3 and wavelengths of 318, 396, and 422 nm were used to measure the Ca 2+ concentrat ion. The relative sta ndar d deviation between replicate measurements of each s ample at each wavelength was take n as the error in the concentration measurements. Then, an average concentration was found across the three wavelengths for each sample. Using th e known volume of spi ke s olution added to the system and the measured c oncent rations for the spike solution an d the diluted sample from the water system, a volume of 50.33(1) L was found for the first measurement and 49.05(1) L for the second measurement. With the known water volu mes that were removed from the system through these measurements and during the irrad i ation, an initial average volume of 52.8(6) L was found. Additionally, multiplication factors for water samples 1 - 4 were found to be 211.1(4), 20 6.1(4), 201.7(4), and 190 (2) respectively. The factors for water sample s 1 - 3 were used in the production rate c alculations, while the factor for water sample 4 was used to find a collection efficiency of 47 Ca from the water system. Geometry Correction Me asurements A point s ourc e of the radionuclides eluted from the collect ion re sin bed in Chapter 6 was made by p utting 0.1 mL of this eluate into a microcentrifuge tube. This sample was measured at 25 and 50 cm from the detector face. The contents of the tu be were then transfer red completely (transfer verified with gamma spect roscop y measurements of the tube) to a 2 50 mL volume of water in the same plastic container used for measurements of the water samples for 247 the production rate measurement. This sample c ontained the same act ivit y as the point source so any differences in th e meas ured activity other than a decay c orrection was attributed to the difference in geometry and the larger volume of water that could attenuate the gamma rays before they escape the sample. Measurements of both the point source and the 250 mL water vol ume we re compared at 25 and 50 cm to fi n d the geometry correction factors. These factors demonstrated a general trend of decreasing with increasing gamma - ray energies. A linear fit was determined for the ge omet ric correction factors with errors from counti ng sta tistics at each distance from the detector. The fit was used to determine correction factors for each gamma ray, including those from shorter - lived radionuclides, such as 44,45 K, 34m,38,39 Cl, and 38 S, and the longer - lived 46 Sc which were not detected in th e sample used but did require a c o rrection factor for other measured samples. The errors associated with each factor were found by propagating the uncertainty in the slope and y - i ntercept determined w ith the linear fit. The measured correction factor s and the linear trend found for each d i stance are shown in Figure 5.2. In practice, the geometry correction factors were used to multiply the measured activities by each gamma line in the water samples use d to calculate the production rates. This correcti on was necessary as the raw activity me a surements from the water samples are 5 - 20% lower than the true activity present and would have produced an underestimated production rate for all radionuclides. 248 Figu re D 1 : Linear Trend s Used to Interpolate Geometry Correcti o n Factors The correction factors for 25 cm (a) and 50 cm (b) from the detector face are given with a linear fit to these data points. 249 REFERE NCES 250 REFERENCES [1] M. Heron, Natl. Vita l Stat. Reports 66 , ( 2017). [2] M . B. Sporn, Lancet 347 , 1377 (1996). [3] C. Müller, Molecules 18 , 5005 (2013). [4] K. Siwowska, R. Schmid, S. Cohrs, R. Schibli, and C. Müller, Ph armac euti cals 10 , 72 (2017). [5] C. Müller, M. Bunka, S. Haller, U. Köster, V . Groehn, P. Bern hardt, N. Van De r Meulen, A. Türler, and R. Schibli, J. Nucl. Med. 55 , 1658 (2014). [6] M. Fani, L. Del Pozzo, K. Abiraj, R. Mansi, M. L. Tamma, R. Cescato, B. Wa ser , W. A. Weber, J. C. Reubi, and H. R. Maecke, J. Nucl. Med. 52 , 1110 (2011 ). [7] S. C. Sr ivastava, Semin Nu cl Med 42 , 151 (2012). [8] T. W. Burrows, Nucl. Data Sheets 108 , 923 (2007). [9] J. 36 , 19 02 (1995). [10] Haller, U. K öster, B. Ponsard, R. Schibli, A. Türler, and N. P. van der Meulen, EJNMMI Radiopharm. Chem. 2 , 5 (2017) . [11] C. Müller, K. A. Domnanich, C. A. Umbricht, and N. P. Va n Der Meulen, Br. J. Radiol. 91 , (2018). [12] E. Eppard, A. De La Fuente, M. A. Khawar, R. A. Bundsc huh, F. C. Gärtner, B. Kreppel, K. Kopka, M. Essler, and F. Rösch, Theranostics 7 , 4359 (2017). [13] C. Muller, M. Bunka, J. Reber, C. Fischer , K. Zhernosekov, A. Turler, and R. Schibli, J. Nucl. Med. 54 , 2168 (2013). [14] B. Singh and J. Chen, Nucl. Data Sheets 126 , 1 (2015). [15] J. Chen, B. Singh, and J. A. Cameron, Nucl. Data Sheets 112 , 2357 (2011). [16] E. A. McCutchan, Nucl. Data Sheet s 11 3 , 1735 (2012). [17] F. G. Kondev, Nucl. Data Sheets 159 , 1 (2019). [18] R. P. B aum, rmeulen, S. Gnesin, U. Köst er, K. Johnston, D. 251 Müller, S. Senftleben, H. R. Kulkarni, A. Türler, R. Schibli, J. O. Prior, N. P. van der M eule n, C. Müller, J. Stanja, F. Wienholtz, and K. Zuber, Dalt. Trans. 54 , 2121 (2017). [1 9] R. Misiak, R. Walczak, B. W elski, and A. Bilewicz, J. Radioanal. Nucl. Chem. 313 , 429 (2017). [20] M. U. Khandaker, K. Kim, M. W. Lee, K. S. Kim, G. N. Kim, Y. S. Cho, and Y. O. Lee, Appl. Radiat. Isot. 67 , 1348 (2009). [21] L. F. Mausner, K. L. Kolsky, V. Joshi, and S. C. Srivastav a, Appl. Radiat. Isot. 49 , 285 (1998). [22] S. Rane, J. T. Harris, and V. N. Starovoitova, Appl. Radiat. Isot. 97 , 188 (2015). [23] M. Yagi and K. Kondo, Int. J. Appl. Radiat. Isot. 28 , 463 (1977). [24] M. Mamtimin, F. Harmo n, an d V. N. Starovoitova, Appl. Radiat. Isot. 102 , 1 (2015). [25] C. S. Loveless, L. L. Radford, S. J. Ferran, S. L. Queern, M. R. Shepherd, and S. E. Lapi, EJNMMI Res. 9 , (2019). [26] D. A. Rotsch, M. A. Brown, J. A. Nolen, T. Brossard, W. F. Henning, S. D. C hemerisov, R. G. Gromov, an d J. Greene, Appl. Radiat. Isot. 131 , 77 (2018). [27] L. K. Peker, Nucl. Data Sheets 68 , 271 (1993). [28] T. W. Burrows, Nucl. Data Sheets 107 , 1747 (2006). [29] O. B. Tarasov and D. Bazin, (2008). [30] E. P. Abel, M. Avilov, V. Ayres, E. Birnbaum, G. Bollen, G. Bonito, T. Bredeweg, H. Clause, A. Couture, J. DeVore, M. Dietrich, P. Ellison, J. Engle, R. Ferrieri, J. Fitzsimmons, M. Friedman, D. Georgobiani, S. Graves, J. Greene, S. Lapi, C. S. Lovel ess, T . Mastren, C. M artinez - Gom ez, S. McGuinness, W. Mittig, D. Morrissey, G. Peaslee, F. Pellemoine, J. D. Robertson, N. Scielzo, M. Scott, G. Severin, D. Shaughnessy, J. Shusterman, J. Singh, M. Stoyer, L. Sutherlin, A. Visser, and J. Wilkinson, J. Phys . G Nu cl. Part. Phys. 46 , 100501 (2019). [31] D. J. Morrissey, B. M. Sherrill, M. Steiner, A. Stolz, and I. Wiedenhoever, Nucl. Instruments Methods Phys. Res. Sect. B Beam Interact. with Mater. Atoms 204 , 90 (2003). [32] A. Pen, T. Mastren, G. F. Peaslee, K. Pet rasky, P. A. Deyoung, D. J . Morrissey, and S. E. Lapi, Nucl. Instruments Methods Phys. Res. Sect. A Accel. Spectrometers, Detect. Assoc. Equip. 747 , 62 (2014). 252 [33] T. Mastren, A. Pen, G. F. Peaslee, N. Wozniak, S. Loveless, S. Essenmacher, L. G. Sob otka, D. J. Morris sey, and S. E. Lapi, Sci. Rep. 4 , 6706 (2014). [34] T. Mastren, A. Pen, S. Loveless, B. V. Marquez, E. Bollinger, B. Marois, N. Hubley, K. Brown, D. J. Morrissey, G. F. Peaslee, and S. E. Lapi, Anal. Chem. 87 , 10323 (2015). [35] C. S. Lov eless, B. E. Mar ois, S. J. Ferra n, J. T. Wilkinson, L. Sutherlin, G. Severin, J. A. Shusterman, N. D. Scielzo, M. A. Stoyer, D. J. Morrissey, J. D. Robertson, G. F. Peaslee, and S. E. Lapi, Appl. Radiat. Isot. 157 , 109023 (2020). [36] K. A. Domnan ich, E. P . Abel, H. K. C lause, C. Kalman, W. Walker, and G. W. Severin, Nucl. Instruments Methods Phys. Res. Sect. A Accel. Spectrometers, Detect. Assoc. Equip. 959 , 163526 (2020). [37] E. P. Abel, H. K. Clause, and G. W. Severin, Appl. Radiat. Isot. 158 , 109049 (2 020). [38] T. Mastren, A. Pen, G . F. Peaslee, N. Wozniak, S. Loveless, S. Essenmacher, L. G. Sobotka, D. J. Morrissey, and S. E. Lapi, Sci. Rep. 4 , 1 (2014). [39] M. Avilov, A. Aaron, A. Amroussia, W. Bergez, C. Boehlert, T. Burgess, A. Carroll, C . Colin, F. Durantel, P. Ferrante, T. Fou rmeau, V. Graves, C. Grygiel, J. Kramer, W. Mittig, I. Monnet, H. Patel, F. Pellemoine, R. Ronningen, and M. Schein, Nucl. Instruments Methods Phys. Res. Sect. B Beam Interact. with Mater. Atoms 376 , 24 (2016). [40] E. P. Ab el, K. Domn anich, C. Kalman, W. Walker, J. W. Engle, T. E. Barnhart, and G. Severin, Nucl. Instruments Methods Phys. Res. Sect. B Beam Interact. with Mater. Atoms 478 , (2020). [41] E. P. Abel, K. Domnanich, H. K. Clause, C. Kalman, W. Walker, J. A . Shuster man, J. Gr eene, M. Gott, and G. W. Severin, ACS Omega (2020). [42] D. W. McKee, J. Catal. 14 , 355 (1969). [43] A. Gavron, Phys. Rev. C 21 , 230 (1980). [44] D. B. O.B. Tarasov, Nucl. Instruments Methods Phys. Res. B 204 , 174 (2003). [45] V. Kanike, J. meesu ngnoen , and J. P. Jay - Gerin, Aus tin J. Nucl. Med. Radiother. 2 , 1011 (2015). [46] V. Kanike, J. Meesungnoen, and J. - P. Jay - Gerin, RSC Adv. 5 , 43361 (2015). [47] J. Meesungnoen and J. P. Jay - Gerin, J. Phys. Chem. A 109 , 6406 (2005). 253 [48] B. Pastina and J. A . L aVerne, J. Phys. Chem. A 103 , 1592 (1999). [49] J. F. Ziegler, M. D. Ziegler, and J. P. Biersack, Nucl. Instruments Methods Phys. Res. Sect. B Beam Interact. with Mater. Atoms 268 , 1818 (2010). [50] M. S. Basunia, Nucl. Data Sheets 114 , 1189 (2 013). [51 ] T. W. Burrows, Nucl. Data Shee ts 108 , 923 (2007). [52] Y. Dong and H. Junde, Nucl. Data Sheets 128 , 185 (2015). [53] R. B. Firestone, Nucl. Data Sheets 108 , 2319 (2007). [54] N. Nica and B. Singh, Nucl. Data Sheets 113 , 1563 (2012). [55] D. R. Tilley , C. M. Cheves, J. L. Godwin, G. M. Hale, H. M. Hofmann, J. H. Kelley, C. G. Sheu, and H. R. Weller, Nucl. Phys. A 708 , 3 (2002). [56] D. R. Tilley, H. R. Weller, C. M. Cheves, and R. M. Chasteler, Nucl. Phys. A 595 , 1 (1995). [57] N. E. Bibler, J. Phys. Chem. 79 , 1991 (1975). [58] J. A. LaV erne, J. Phys. Chem. 92 , 2808 (1988). [59] J. A. LaVerne, in Charg. Part. Phot. Interact. with Matt er Chem. Physicochem. Biol. Consequences with Appl. (n.d.). [60] F. Nelson, T. Murase, and K. A. Kraus, J. Chromatog r. A 1317 8 , 505 (1964). [61] K. A. Kraus and F. Nelson, in Proc. Int. Conf. Peac. Uses At. Energy Nucl. Chem. Eff. Irradiat. (1956), pp. 113 125. [62] N. Takeno, Atlas of Eh - PH Diagrams: Intercomparison of Thermodynamic Databases (2005). [63] C. S. Love less, B. E. Marois, S. J. Ferran, J. T. Wilk inson, L. Sutherlin, G. Severin, J. A. Shusterman, N. D. Scielzo, M. A. Stoyer, D. J. Morrissey, J. D. Robertson, G. F. Peaslee, and S. E. Lapi, Appl. Radiat. Isot. 157 , 109023 (2020). [64] T. Sato, Y. Iwamoto, S. Hashimoto, T. Ogawa, T. Furuta, S. Abe, T. Kai, P. Tsai, N. Matsuda, H. Iwase, N. Shigyo, L. Sihver, and K. Niita, J. Nucl. Sci. Technol. 55 , 684 (2018). [65] M. Shamsuzzoha Basunia, Nucl. Data Sheets 112 , 1875 (2011). 254 [66] B. Singh and J. A. Cameron, Nucl. Data She ets 92 , 1 (2001). [67] T. W. Bur rows, Nucl. Data Sheets 109 , 171 (2008). [68] K. Neubauer and L. Thompson, Spectroscopy 26 , 24 (201 1). [69] R. (Eds. . Qaim, S.M., Tárkayáni, F., Capote, Iaea - Tecdoc - 1211 292 (2001). [70] E. P. Abel, H. K. Clause, and G. W. Severin, Appl. Radiat. Isot. 158 , (2020). [71] T. E. Gangwer, M. Goldstein, and K. K. S. Pillay, (1977). [72] Y. Sugo, Y. Izumi, Y. Yoshida, S. Nishijima, Y. Sasaki, T. Kimura, T. Sekine, and H. Kudo, Radiat. Phys. Chem. 76 , 794 (2007) . [73] E. P. Horwitz, D. R. McAlister, A. H. Bond, an d J. E. Barrans, Solvent Extr. Ion Exch. 23 , 319 (2005). [74] S. S. Raiman, A. Flick, O. Toader, P. Wang, N. A. Samad, Z. Jiao, and G. S. Was, J. Nucl. Mater. 451 , 40 (2014). [75] P. Wang and G. S. Wa s, J. Mater. Res. 30 , 13 35 (2015). [76] W. Jimin and H. Xiaolong, Nucl. Data Sheets 144 , 1 (2017). [77] F. T. Tárkányi, A. V. Ignatyuk , A. Hermanne, R. Capote, B. V. Carlson, J. W. Engle, M. A. Kellett, T. Kibedi, G. N. Kim, F. G. Kondev, M. Hussain, O. Lebeda, A. Luca, Y. Nagai, H. Naik, A. L. Nichols, F. M. N ortier, S. V. Suryanarayana, S. Takács, and M. Verpelli, Recommended Nuclear D ata for Medical Radioisotope Production: Diagnostic Gamma Emitters (2019). [78] J. M. Alexander and D. H. Sisson, Physi cal 128 , 2288 (1962). [79] M. de Jong, R. Valkema, F. Jamar , L. K. Kvols, W. A. P. Breeman, C. Smith, S. Pauwels, and E. P. Krenning, Sem in. Nucl. Med. XXXII , 133 (2002). [80] F. W. E. Strelow, 56 , 1053 (1984). [81] D. Reher, H. H. Hansen, R. Vaninbrouk x, M. J. Woods, C. E. Grant, S . E. M. Lucas, J. Bouchard, J. M orel, and R. Vatin, Appl. Radiat. Isot. 37 , 973 (1986). [82] J. W. T. Meadows and V. A. Mode, J. Inorg. Nucl. Chem. 30 , 361 (1968). 255 [83] H. Mommsen, I. Perlman, and J. Yellin, Nucl. Instrument s Methods 177 , 545 (1980). [84] M. B. Lewis, Nucl. Data Sheets 3 13 (1970). [85] L. Burkinshaw, D. H. Marshall, and C. B. Oxby , Int. J. Appl. Radiat. Isot. 20 , 393 (1969). [86] M. S. Freedman, F. T. Porter, and F. Wagner, Phys. Rev. 152 , 1005 (1966). [8 7] R. E. Wood, J. M. Palms, and P. V . Rao, Nucl. Physics, Sect. A 12 6 , 300 (1969). [88] H. J. Fischbeck, Phys. Rev. 173 , 1078 (1968).