in partial fulfillment of the requirements for the degree of Doctor of Philosophy (Ph.D)

Transcription

1

2 Investigations of Nuclear Forensic Signatures in Uranium Bearing Materials A dissertation submitted to the Graduate School of the University of Cincinnati in partial fulfillment of the requirements for the degree of Doctor of Philosophy (Ph.D) In the Department of Chemistry Of the McMicken College of Arts and Sciences By Lisa Ann Meyers B.S. Ohio Northern University 2009 August 2013 Committee Chairs: Thomas Beck, Ph.D. Apryll Stalcup, Ph.D. i

3 Abstract Nuclear forensics is a multidisciplinary science that uses a variety of analytical methods and tools to investigate the physical, chemical, elemental, and isotopic characteristics of nuclear and radiological material. A collection of these characteristics is called signatures that aids in determining how, where and when the material was manufactured. Radiological chronometry (i.e., age dating) is an important tool in nuclear forensics that uses several methods to determine the length of time that has elapsed since a material was last purified. For example, the age of a uranium-bearing material is determined by measuring the ingrowth of 230 Th from its parent, 234 U. A piece of scrap uranium metal bar buried in the dirt floor of an old, abandoned metal rolling mill was analyzed using multi-collector inductively coupled plasma mass spectroscopy (MC-ICP- MS). The mill rolled uranium rods in the 1940s and 1950s. The age of the metal bar was determined to be 61 years at the time of analysis using the 230 Th/ 234 U chronometer, which corresponds to a purification date of July 1950 ± 1.5 years. Radiochronometry was determined for three different types of uranium metal samples. The affects of different etching procedures were evaluated to determine whether etching procedures affect radiological age. The sample treated with a rigorous etching procedure (concentrated HNO 3 ) had exhibited the most reliable radiological age while less rigorous etching (8M HNO 3 ) yields a radiological age from 15 years to hundreds of years older than the known age. Any excess thorium on the surface of a uranium metal sample presents a bias in age determination and the sample will appear older than the true age. Although this research demonstrates the need for rigorous surface etching, a bias in ii

4 the radiological age could have arisen if the uranium in the metal was heterogeneously distributed. The uranium isotopic composition of radiological samples can reveal whether the uranium is natural, depleted, or enriched and whether the material has ever been subjected to neutron irradiation or reprocessing. The signatures contained in samples of dirt collected at two different uranium metal processing facilities in the United States were evaluated to determine uranium isotopic composition and compare results with processes that were conducted at these sites. One site refined uranium and fabricated uranium metal ingots for fuel and targets and the other site rolled hot forged uranium and other metals into dimensional rods. Unique signatures were found that are consistent with the activities and processes conducted at each facility and establish confidence in using these characteristics to reveal the provenance of other materials that exhibit similar signatures. In this research, nuclear forensics signatures of a variety of different types of uranium bearing samples, including metals, soils, and ore deposits were determined. Radiochronometry, uranium, thorium, and plutonium isotopic analysis, and elemental composition of these materials were all analyzed. These characteristics, when evaluated alone or in combination, become signatures that may reveal how and when the material was fabricated. iii

5 iv

6 Acknowledgments None of my accomplishments would have been possible without a loving, strong, supporting, and encouraging group of people close by me. I would first like to thank my PhD committee (Dr. Beck, Dr. Ridgway, Dr. Connick, Dr. Spitz, and Dr. Stalcup) and the Department of Chemistry at UC for all of their support and guidance throughout this process. I would like to especially thank my co-advisors, Dr. Apryll Stalcup and Dr. Henry Spitz, who has helped me grow as both a scientist and a person. They have opened up many unique opportunities in my graduate school career that I wouldn t have otherwise. I would like to thank Dr. LaMont and Dr. Glover for their mentorship and support. I would like to thank the past and present Stalcup and Spitz group for all their support and encouragement. I would like to thank Dr. Ross Williams for his expertise and assistance for my dissertation research at Lawrence Livermore National Laboratory, along with Dr. Kim Knight, Dr. Sarah Roberts, and Rachel Lindvall who helped me with instrumentation operations and/or ran analyses on my samples. I would like to thank Kim Carey for answering my random questions, working hard at scheduling my seminars, as well as providing help with anything I needed. I strongly believe that the department of chemistry wouldn t be able to function without her. I would like to thank my undergraduate research advisor Dr. Celius and his wife Rachael, who always had faith in my future and accomplishments and have always been very supporting even throughout my time in graduate school. Finally, I would like to thank my family and friends for their daily love and support. This dissertation would not have been possible without all of you. Thank you. This work wouldn t have been able to be accomplished without the help of funding from various places. I would like to thank the U.S. Department of Energy s Nuclear Materials Information Program for funding this research. This work was part performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344, support by the U.S. Department of Homeland Security under Grant Award Number 2012-DN-130-NF , and the Nuclear Forensics Graduate Fellowship Program, which is sponsored by the U.S. Department of Homeland Security, Domestic Nuclear Detection Office and the U.S. Department of Defense, Defense Threat Reduction Agency. The views and conclusions contained in this document are those of the authors and should not be interpreted as necessarily representing the official policies, either expressed or implied, of the U.S. Department of Homeland Security. v

7 Table of Contents Abstract... ii Acknowledgements...v List of Tables... viii List of Figures... ix Chapter 1: Nuclear Forensics Signatures and Techniques Introduction to Nuclear Forensics Nuclear Forensic Signatures Uranium and Thorium Isotopic Signatures Plutonium Isotopic Signatures Radiochronometry Instrumentation Techniques Non-destructive Surface Characterization Techniques Scanning Electron Microscopy (SEM) and Electron Dispersive Spectroscopy (EDS) X-ray Diffraction (XRD) Destructive Elemental Techniques Inductive Coupled Plasma Mass Spectrometry (ICP-MS) Multi-collector ICP-MS (MC-ICP-MS) Isotope Dilution Mass Spectrometry (IDMS) Actinide Separations Research Focus References...25 Chapter 2: Radiochronological Age of a Uranium Metal Sample from an Abandoned Facility Introduction Experimental Methods...31 vi

8 2.2.2 Sample preparation of soil Radiochemical separation and purification chromatography of soil samples Sample preparation of uranium metal Radiochemical separation and purification chromatography of metal thorium fractions Instrumentation Results and Discussion Conclusions References...40 Chapter 3: Confidence in Radiological Chronometry of Uranium Metal Samples Introduction Experimental Materials Sample preparation of uranium metal Radiochemical separation and purification chromatography of thorium metal fractions Instrumentation Results and Discussion Conclusions References...52 Chapter 4: Uranium Isotopic Signature Analysis of Collected Dirt Samples at Former Metal Rolling Mills Introduction Experimental Materials Methods Sample Preparation for U/Th Analysis U/Th Radiochemical separation and purification chromatography Sample Preparation for Pu Analysis Pu Radiochemical separation and purification chromatography Instrumentation...61 vii

9 4.3 Results Discussion Conclusions References...66 Chapter 5: Elemental and Auxiliary Nuclear Forensic Signatures in a Variety of Uranium Bearing Materials Introduction Experimental Materials Surface Analysis Techniques SEM/EDS XRD Sample Preparation Methods Uranium and thorium sample preparation for ore samples Uranium metal and soil sample preparation Elemental and rare earth elemental analysis sample preparation Destructive Instrumentation Techniques MC-ICP-MS Quad-ICP-MS Results and Discussion Uranium metal surface comparison Uranium metal elemental analysis Ore sample surface analysis Uranium Ore Elemental Analysis Soil, Uranium Metal, and Ore Sample Elemental Comparison Rare Earth Elemental Analysis Conclusions References...96 viii

10 Chapter 6: Conclusions and Future Directions Conclusions Future Directions Development of a 230 Th/ 234 U Age Dating Standard Reference Material using CRM-112A Investigations of nuclear forensic signatures using different matrices (e.g. rusty metal) at former metal fabrication facilities Investigations of Signatures at Multiple Former Uranium Metal Rolling Facilities Using Rare Earth Element Signatures for Nuclear Forensic Purposes References ix

11 List of Tables 1.1: Naturally occurring uranium isotope abundances in varying enrichments of uranium : Uranium and thorium negatively charged complexes in the +4 and +6 oxidation state : Distribution coefficients of actinides on anion exchange resins in pure HCl solutions : Distribution coefficients of actinides on anion exchange resins in pure HNO 3 solutions : List of extraction chromatography resin types available commercially and the actinides it can selectively separate : The organic stationary phase for UTEVA and TEVA extraction chromatography resin : Uranium isotopic composition in samples : Total uranium concentration in samples : Predicted model age of each uranium metal sample : Thorium isotopic and concentration for each uranium metal sample : Uranium and plutonium concentrations in soil samples : Elemental composition of uranium metal samples by ICP-MS : Uranium isotopic composition and concentration in uranium metal samples : Uranium concentration comparison between ore samples : Elemental composition of uranium ore samples by ICP-MS : Elemental composition of former metal fabricating soil samples by ICP-MS...90 x

12 List of Figures 1.1: First half of uranium and thorium decay chains : A simple schematic of a scanning electron microscope : Illustration of specific signals generated from the interactions between the incident beam and sample and what regions within the sample these signals can be generated : A simple schematic illustrating the basic functions of XRD : Schematic of an ICP-MS instrument : (Top): A schematic of the Nu Plasma HR multi-collector ICP-MS. (Bottom): A close-up of the optical and detector set-up schematic : UTEVA resin stationary phase log distribution coefficient for actinides plotted against the concentration of HNO 3 (left) and HCl (right) : TEVA resin stationary phase log distribution coefficient for actinides plotted against the concentration of HNO 3 (left) and HCl (right) : 238 U decay chain series displays the decay of 234 U into 230 Th : Soil and uranium metal sample 235 U/ 238 U ratio comparison : Age determination comparison of each sample. Error bars are expanded uncertainties (k=3) : Photographs of the uranium metal samples: Uranium rod (U-rod) (left), F-element Solid (middle), and F-element Drillings (right) : Age determination comparison of each sample. Error bars are expanded uncertainties (k=3) : McCulloch and LLNL age dating of CRM112-A comparison : 235 U/ 238 U ratio comparison for soil samples (Solid line indicates ratio for natural uranium) : 236 U/ 235 U vs. 238 U/ 235 U ratio comparison plot for soil samples : 234 U/ 235 U vs. 238 U/ 235 U ratio comparison plot for soil samples : Rare earth element concentrations plotted against Z number with and without normalization : SEM images of U-rod-1 oxidized side (left) demonstrates the presence of trace impurities and the shiny metallic side shows crystal formations indicated by black circle (right)...76 xi

13 5.3: SEM images of uranium metal samples for comparison: A. U-rod-1, B. U-rod-2, C. U-rod-3, D. F-element Solid, E. F-element Drillings 1, F. F-element Drillings : Elemental radar plots for element comparison in uranium metal samples. The numbers indicate normalized elemental concentrations (µg/g ) for each sample : Uranium isotopic ratio correlation plot demonstrating F-element Drillings 1 and 2 have significant 236 U signature : Uranium isotopic ratio correlation plot demonstrating F-element Drillings 1 and 2 have significant 234 U signature : SEM images of Western ore (left) and Pitchblende ore (right) : XRD spectrum of western US ore sample : XRD spectrum of pitchblende ore sample : Uranium isotopic ratio correlation plot indication Western US ore has a higher 234 U signature : Elemental radar plots for element comparison in uranium ore samples. The numbers indicate normalized elemental concentrations (mg/g) for each sample : Elemental radar plots for element comparison in former metal rolling facilities soil samples. The numbers indicate normalized elemental concentrations (mg/g ) for each sample : Venn Diagram of common elemental make-up between sample types : REE log chondrite normalization plots for uranium metals, ores, and F-waste samples : Close up REE log chondrite normalization plots for western ore, pitchblende ore, and F-waste samples...95 xii

14 Chapter 1: Nuclear Forensics Signatures and Techniques 1.1 Introduction to Nuclear Forensics Nuclear forensics is a multidisciplinary field of science involving chemistry, physics, and engineering that was developed as an outcome of a national effort to evaluate suspicious materials that were obtained through covert surveillance or monitoring practices. Many different physical and chemical processes are utilized to analyze isotopic concentrations of significant nuclides, trace element content, microstructure, and the elapsed time since a material last purified. This last characteristic is considered as the radiological age of a substance. Results of these analyses determine a unique signature for a material that can aid in revealing its provenance, origin, a date of fabrication or purification, how the material was made, and (perhaps) the route used to transport material from its origin to the point of interception [1, 2]. Developing nuclear forensic signatures and determining what types of information are important to search for can help distinguish one radiological or nuclear material from another and can also be used to identify the process that was initially used to create these types of materials. Both non-destructive analytical methods (e.g., scanning electron microscopy, X-ray fluorescence, and alpha/gamma spectroscopy) and destructive analytical techniques (i.e., mass spectrometry and analytical separations) are used to analyze many types of nuclear or radiological material for forensic signatures [3]. The most important of these analytical methods will be discussed in the following sections. 1

15 1.2 Nuclear Forensic Signatures Uranium and Thorium Isotopic Signatures Uranium isotopic composition of radiological or nuclear material can reveal information about the history of the material, how it was used, or the type of processes that were involved in fabricating the material. It also can reveal whether the uranium is natural, depleted, or enriched and whether the material has ever been subjected to neutron irradiation and subsequently reprocessed. Uranium isotopic analysis has been used in nuclear forensics to characterize environmental samples [4] and seized materials [5] including samples of spent nuclear fuel and here-to-fore unknown activities associated with concealed nuclear facilities [6-10]. Isotopic concentrations of uranium have also been reported for samples collected following the Chernobyl accident [11,12], the Balkan s conflict [13], and incidents associated with weapons development in the United Kingdom [14]. Retrospective analysis of the uranium isotopic composition in these samples provides a comprehensive set of data to relate isotopic signatures to processes and activities associated with the production and use of nuclear or radiological materials. There are three naturally occurring uranium isotopes with long half-lives, viz., 238 U, 235 U, and 234 U. The isotopic composition of natural uranium is ± % 238 U, ± % 235 U, and ± % 234 U by mass [15]. Whether a sample of uranium is natural, depleted, or enriched in the 235 U isotope can be determined by evaluating the 235 U/ 238 U mass ratio. 235 U is important for nuclear reactors and weapons because of its fissile characteristics. The most common value for the natural 235 U/ 238 U mass ratio is [16-18]. Table 1.1 demonstrates the abundances of the naturally occurring uranium isotopes in natural, depleted, and enriched uranium materials [19]. 2

16 Table 1.1. Naturally occurring uranium isotope abundances in varying enrichments of uranium [19] Uranium used in nuclear fuel and weapons generally require enrichment of the 235 U isotope significantly above that in natural uranium. The presence of 236 U in a sample is an important isotopic signature that suggests the uranium may have been reprocessed after exposure in a nuclear reactor [11] as 236 U is created from the neutron bombardment of 235 U. 232 U and 233 U are isotopes that also indicate material have been previously irradiated. 232 U is produced as follow: 235 U decays into the short-lived isotope, 231 Th, which, is then decayed into 231 Pa. When 231 Pa is exposed to a neutron flux, 231 Pa accepts a neutron and becomes 232 Pa. 232 Pa then quickly decays into 232 U. 233 U is produced from the neutron absorption of 232 U or from a series of decays following the neutron bombardment of 232 Th in a reactor [2]. 232 U is also an impurity from irradiation of 232 Th when producing 233 U. The first half of the uranium and thorium naturally decay chains are shown in Figure 1.1. There are six naturally occurring thorium isotopes that can be detected in this figure: 234 Th, 230 Th, 231 Th, 227 Th, 232 Th, and 228 Th. The two most stable thorium isotopes are 232 Th and 230 Th indicated by their long half-lives in Figure Th is the beginning of its own decay series. With the presence of an abundance of 232 Th and little to no other thorium isotopes suggests that uranium-bearing materials have not been recently purified. 3

17 Figure 1.1. First half of uranium and thorium decay chains from Plutonium Isotopic Signatures Plutonium isotopes are primarily related to the burn-up of uranium fuel in reactors and their abundances can depend on the type of reactor in which they are generated and the operating history of the reactor. The most common mechanism of generating plutonium is when 238 U absorbs a neutron creating 239 U. 239 U decays into 239 Np, which decays into 238 Pu. Other Pu isotopes (i.e. 240 Pu, 241 Pu, 242 Pu, etc.) are primarily produced via subsequent neutron absorptions of 239 Pu [20]. The longer uranium fuel is irradiated in a reactor, the larger the quantity of plutonium is produced. Therefore, plutonium isotopic composition changes significantly with the age of the material and length of irradiation. The buildup of 241 Am can be used to determine the time that has elapsed since that material was last purified [21]. 239 Pu is a fissile material and is popularly used in weapons 4

18 for this reason. The 239 Pu/ 240 Pu ratio determines the length of time the material was irradiated in a reactor and the purpose of its usage. A larger amount of 239 Pu is generated with short irradiation exposure. Therefore, weapons grade plutonium contains less than 6% 240 Pu [20] Radiochronometry Radiochronometry (i.e., age dating) determines the time elapsed since purification of radioactive or nuclear material based upon the in-growth of a daughter nuclide from the decay of its parent isotope in a non-disturbed environment. Three chronometers used to determine the radiological age of uranium and thorium bearing material include: 230 Th/ 234 U [10, 22-26], 228 Th/ 232 Th [27], and 231 Pa/ 235 U [28-29]. The most common chronometer used of these three is 230 Th/ 234 U. Use of the 230 Th/ 234 U chronometer assumes that all contaminants were removed when the material was last purified and requires complete separation of thorium from uranium and the removal of oxidation and residual contamination on the surface of the uranium metal sample prior to age determination [27]. Common chronometers used to age date plutonium bearing materials include: 238 Pu/ 234 U, 239 Pu/ 235 U, 240 Pu/ 236 U and 241 Pu/ 241 Am [21]. These chronometers must also assume similar constraints to uranium/thorium chronometers. The process of radiochronometry typically employs several analytical methods that may include thermal ionization mass spectrometry (TIMS) [22,30], inductively coupled plasma mass spectrometry (ICP-MS) [6, 8, 23,31-33], gamma spectroscopy [29, 34-35], and isotope dilution alpha spectrometry [7,25] to determine the isotopic composition and radiological age of radiological or nuclear material. 5

19 1.3 Instrumentation Techniques Non-destructive Surface Characterization Techniques Prior to destroying the physical appearance of a sample, scanning electron microscopy (SEM) can be used to determine physical or morphologic information about a particular sample. Electron dispersive spectroscopy (EDS) is usually coupled to a SEM and can provide qualitative or quantitative elemental composition. This is a great method to pre-screen samples for specific characteristics or elemental composition. To determine physical phases of a sample, a semi-destructive X-ray diffraction (XRD) technique, can be used Scanning Electron Microscopy (SEM) and Electron Dispersive Spectroscopy (EDS) Scanning Electron Microscopy (SEM) is a very versatile technique used for the characterization and analysis of microstructure morphology on an array of conducting or semi-conducting materials from the formation of a 3-D image. A schematic of SEM is displayed in Figure 1.2 [36]. An electron gun is used to emit a beam of electrons that travel through a series of focusing apparatuses made up of lenses and deflection coils in order to focus the electron beam towards the sample located on a platform. The electrons from the beam interact with the sample and give off signals such as secondary electrons, backscatter electrons, characteristic x-rays, and other photons of various energies. These signals are obtained from specific emission points within the sample that can be used to examine the sample s surface topography, crystallography, and composition. A variety of detectors are placed in specific geometries corresponding to the position of the beam course and sample location to detect these signals given off from the interaction between 6

20 the electron beam and the sample. Figure 1.3 illustrates the regions of a sample in which the specific signals can be detected [37]. Figure 1.2: A simple schematic of a scanning electron microscope. [36] 7

21 Figure 1.3: Illustration of specific signals generated from the interactions between the incident beam and sample and what regions within the sample these signals can be generated. [37] The two most popular signals used to form a nanometer to micrometer scale image of a sample are backscattered electrons and secondary electrons. Incident electrons that scatter through an angle of more than 90 degrees off of the sample are called backscattered electrons (BSE). These electrons possess energies greater than 50 electron volts (ev). BSE are used for the examination of topographic contrast and compositional characteristics in a sample. When the primary beam strikes the sample surface causing the sample s atoms to ionize, loose bound electrons may be emitted generating secondary electrons (SE). Secondary electrons are known to possess energies of less than 50 ev and are mainly used for topographic contrast and the visualization of surface texture of a sample [37-38]. The analysis of characteristic x-ray signals is mainly used to determine the chemical composition of a sample. The main x-ray spectrometer used with SEM is electron-dispersive spectroscopy (EDS). EDS excels at obtaining rapid qualitative analysis since it can acquire spectra covering a large range of energies at one time [39]. This is a great technique for pre-screening samples for elemental composition as a nondestructive method. This technique can also be used for quantitative chemical analysis and compositional mapping of samples. Majority of the SEMs that are sold today are equipped with EDS capabilities [38] X-ray Diffraction (XRD) 8

22 X-ray diffraction (XRD) is a non-destructive analytical technique used to identify and quantify crystalline phases and perform structural analysis in a sample. Single or multiple crystalline phases may be present and should be able to be identified. A simple schematic illustrating the basic functions of XRD is displayed in Figure 1.4 [40]. X-rays are generated using an x-ray tube. The generated x-rays interact with electrons within the sample and emit diffractive characteristic x-rays. The diffractive x-rays go through several slits before striking the detector, which registers the intensity of the x-ray at an angle 2θ with respect to the sample plane. On the left side of Figure 1.4, the parallel lines represent the planes of atoms with a distance of d hkl defined by the miller indices. The incident and reflection beam displayed follows the Bragg Diffraction condition, in which the incident beam angle is equal to the angle of reflection at certain characteristic angles [41-42]. Figure 1.4: A simple schematic illustrating the basic functions of XRD. [40] Each phase within the sample gives rise to a specific diffraction pattern of rings based upon the lattice spacing and this information appear as sharp peaks in a spectrum with intensity plotted against the reflection angle. Matching the diffraction patterns to 9

23 databases identifies the phases inside the sample [43]. Amphorous phases, such as glass, will mask weak reflections resulting in producing a broad background shape as the diffraction pattern [41]. Surface roughness due to coarse grains may affect the relative intensities of the diffraction pattern as well [42] Destructive Elemental Techniques Performing destructive analysis on nuclear forensic samples can provide a variety of information obtained with low detection limits, high accuracy, and high precision techniques. Inductive coupled plasma mass spectrometry (ICP-MS) is a technique used to obtain elemental composition of samples. Techniques such as multi-collector (MC)-ICP- MS and isotope dilution mass spectrometry are advanced techniques that can be used for improved sensitivity, lower detection limits, and higher accuracy and precision than what ICP-MS alone can provide. This type of instrumentation and methods can also provide additional information, such as isotopic ratios and concentrations. Prior to destructive analysis, actinide separations are needed for the removal of interferences and isobars Inductive Coupled Plasma Mass Spectrometry (ICP-MS) ICP-MS is one of the most widely used analytical techniques for elemental analysis because of its low detection limits, good precision and accuracy, and its high degree of sensitivity. A schematic of the ICP-MS system is displayed in Figure 1.5 [44]. Inductive coupled plasma mass spectrometry (ICP-MS) is a destructive analytical technique used for elemental analysis. A more complex version of this, multi-collector ICP-MS (MC-ICP-MS), is a more sensitive method with lower limits of detection that is able to provide both elemental and isotopic information. Isotopic ratios and concentration 10

24 determination via isotope dilution can be accomplished using this instrumentation. MC- ICP-MS and ICP-MS have been widely used in this research to determine elemental and isotopic nuclear forensic signatures. The ICP-MS source is a an ionized argon plasma which can reach a temperature of around 10,000K. A solution of sample is thermally nebulized into small aersol droplets before being introduced into the plasma. The high temperature of the plasma renders the sample components primarily atomized followed by ionization into individual positively charged particles. Once ionized, the particles are extracted from the plasma and are introduced to a quadrupole mass analyzer through a vacuum pumped interface containing two water-cooled nickel metal skimmer cones. The ions are separated and identified based on their mass to charge ratio by traveling through four circular perpendicular pair rods, which are alternating AC and DC potential in opposite pairs of the rods [45-46]. Figure 1.5: Schematic of an ICP-MS instrument. [44] While ICP-MS is a popular technique, it does have its limitations. Interferences are one of the major issues associated with this instrumentation. Isobaric interferences can occur in which two elements may have the same isotope of essentially the same mass (e.g. 232 Th vs. 232 U). This is one of the major issues that need to be addressed when 11

25 analyzing radioisotopes. Chemical separations need to be done prior to analysis in order to overcome this problem. Another type of interference that is also important to address is oxide, hydroxide, or hydride specie interferences (e.g. 233 UH + vs 234 U + ). These types of compounds are formed from interactions with the sample matrix itself or the plasma gas. Polyatomic interferences can happen when elements form molecular in the plasma, sample matrix, or in the atmosphere (e.g. 40 Ar +, 40 ArH + ). This type of interference is less serious compared to isobaric and hydride formation interferences when analyzing radiological or nuclear samples [45-46]. Quadrupole ICP-MS instruments have a mass resolution of about one atomic mass unit, which is sufficient enough for most applications [47]. In the case that a higher resolution is needed, high resolution or sector field mass spectrometers have been used. These type of instruments help reduce or eliminate the interference effects due to mass overlap. Magnetic and electric sectors are used in these types of interments to separate and focus the ions Multi-collector ICP-MS (MC-ICP-MS) MC-ICP-MS has become a popular technique within the last few years as an accurate and precise way of measuring isotopic ratios. The first commercial MC-ICP-MS instrumentation (Plasma 54) was available in 1992 from VG Elemental [48]. Since then, it has continually evolved to one of the most precise techniques for determining isotopic ratios, next to thermal ionization mass spectrometry (TIMS). In addition, for precise 12

26 analysis of isotopic ratios, MC-ICP-MS has also been used for elemental concentration analysis using isotope dilution, which will be discussed more in section A schematic of the Nu Plasma HR multi-collector ICP-MS is displayed in Figure 1.6 [49]. The MC-ICP-MS source and sample introduction is very similar to that of ICP- MS. Once the sample has been atomized and ionized by the ICP plasma, the ions are then accelerated into the expansion chamber between the two cooled nickel skimmer cones, in which most of the sample is lost. Once the ions are able to reach through the chamber and pass by the second cone, the ions are optically focused onto the entrance slit and into the mass analyzer by a series of lenses. The ions of different masses are spatially separated in the focal plane by a electro-magnetic sector mass analyzer using a Nier Johnson doublefocusing geometry which allows for high mass resolution measurements. This geometry occurs when a 90 degree electrostatic sector is located right after the ion source, separates the ions based on their kinetic energy, and used for velocity focusing followed by a 90 degree magnetic sector that separates the ions from one another based on their mass-tocharge ratio [50]. Ions are then detected on multiple Faraday and electron multiplier detectors simultaneously. 13

27 Figure 1.6: (Top): A schematic of the Nu Plasma HR multi-collector ICP-MS. (Bottom): A close-up of the optical and detector set-up schematic. [49] 14

28 The Nu Plasma MC-ICP-MS is equipped with 12 Faraday cup detectors and 3 electron multiplier detectors (also known as ion counting detectors). Faraday cup detectors are robust and accurate at the measurements of ion currents. Electron multiplier detectors are mainly used to measure low ion currents, for lower abundant isotopes. These detectors have signal amplification capability by having the ions strike a dynode that ejects several electrons from each striking ion. The ejected electrons go on to strike additional cascading dynodes, which ejects several secondary electrons. This produces a gain of fold and is then large enough to by detected and counted by pulsecounting electronics [50]. While this type of instrumentation still has challenges to overcome associated with isobaric, hydride, etc. interferences, there are still many advantages. Instrumentation that has the capability to measure several isotopic ion currents simultaneously improves the precision significantly because of the small fluctuations in the ion beam of all ion current since most of the analysis time is spent measuring all ion currents. In this manner, any changes in the conditions of the ion beam will affect all ions in a similar manner. The double-focusing mass analyzer allows for different masses to be specially separated along the focal plane so that multiple detectors can be set up in a specific geometry to be able to measure multiple isotopes at the same time. This sector also allows for high mass resolution (i.e. 10,000) and for flat-top peak shapes. The flat-top peaks allow for the entire ion image to be captured by the detector and that small fluctuations in the position of the ion image on the focal plane doesn t result in variations in the measured ion current intensities. This type of instrumentation also allows for the detection of isotopes covering 15

29 a large dynamic range of isotope ratio abundances with the different types of detectors provided for use in the system [51] Isotope Dilution Mass Spectrometry (IDMS) Isotope Dilution Mass Spectrometry (IDMS) is a technique used in conjunction with ICP-MS for the analysis of a sample with an unknown concentration of a targeted element. A known amount of an isotope-enriched spike compound of the corresponding element or isotope is mixed and equilibrated with the sample. This approach requires the targeted element to have at least two isotopes measurable by ICP-MS. IDMS is a onepoint internal calibration method with the spiked isotope as the calibrant. This technique has multiple advantages in terms of accuracy compared to other multi-point calibration methods. One of the advantages for using this technique is that once the isotope dilution step is completed, the recovery factor doesn t need to be known; this is because the sample and spike will react through the process of analysis in the same manner since the isotope ratios of the unknown target and the spike isotope can be measured. For example, sample loss doesn t affect the analytical results because the sample percentage of spike was lost as well if the spike and sample was fully equilibrated before analysis. Another advantage of IDMS is that it is independent on the sample matrix, for a significant advantage over other calibration approaches. Instrumental signal drift and instability are usually a major problem for other calibration methods, but not for IDMS. The spike and target element are subsequently measured within a short time period that minimal if any drift occurs. If drift causes a problem, a MC-ICP-MS can be used to eliminate the issue. 16

30 Although this approach has many advantages, spectral interferences with either the targeted element or spiked isotope can provide inaccurate results. [52] 1.4 Actinide Separations Chemical separations of actinides are important when using alpha spectroscopy or mass spectrometry for actinide isotope analyses. This allows for any energy or mass interferences to be eliminated (e.g. 232 U vs. 232 Th). Radioactive tracers are often used as an isotope dilution technique in conjunction with chemical separations to perform actinide isotope analyses. When determining the type of tracer to add to a sample, it s important to avoid mass or energy interferences, have a long enough half-life, and consider the cost and availability of the tracer used. 229 Th is a typical tracer used for thorium analyses using both alpha spectroscopy and mass spectrometry. A frequent tracer used for uranium analyses for alpha spectroscopy is 232 U and for mass spectrometry is 233 U. 242 Pu is a common tracer used for plutonium analyses for mass spectrometry. There are a variety of separation methods that could be used to separate one actinide from another including co-precipitation, solvent extraction, ion exchange, and extraction chromatography. The last two will be focused on within this section. Cation and anion exchange separation techniques that can be used to separate actinides. Cation exchange occurs when labile, positively-charged ions, associated with the negatively charged groups along the polymer backbone of the stationary phase, are exchanged by positively charged analytes. Anion exchange occurs when negatively charged counter-ion, associated with positively charged groups along the polymer backbone, exchanges for negative charged analytes. The latter is a more popular 17

31 technique to use for actinides, complexed with either Cl - or NO - 3. In anion exchange chromatography, a typical resin stationary phase used is made up of a styrene divinylbenzene crosslinked, rigid polymer backbone with covalently attached positively charged quaternary ammonium groups with either chloride or nitrate negatively charged counter-ions. The actinide separations are typically performed in either concentrated HCl or HNO 3, which favors formation of polyligand anionic complexes with the metal cations. Retention is dependent on the metal s oxidation states, on the polyligand complex s overall charge, and the metal s distribution coefficient for the particular mobile phase solvent conditions. Uranium s most important and stable oxidation states are in the +4 and +6 states. It is possible for uranium to be in the +2 or +5 oxidation state, but the complexes are weak or unstable. The common complex form for uranium in the +5 or +6 oxidation state + is UO 2 or UO ++ 2, respectively [53]. Common uranium complexes in the +4 and +6 state with an overall -1 or -2 charged nitro complex are shown in Table 1.2. Thorium complexes are the most stable in the +4 oxidation state. It is possible for thorium to be in the +2 or +3 oxidation state, but these complexes are weak or unstable [54]. Common thorium nitro complexes in the +4 oxidation state with an overall -1 or -2 charged complex are also shown in Table

32 Table 1.2: Uranium and thorium negatively charged complexes in the +4 and +6 oxidation state. Hydrochloric acid and nitric acid are the two main mobile phase solvents used in ion exchange chromatography. When the solvent is pure hydrochloric acid at varying molarities (M), the distribution coefficients for varying actinides are displayed in Table 1.3 [55]. Thorium doesn t absorb onto the resin when hydrochloric acid at any concentration is used as a solvent. Uranium absorbs weakly at 4M HCl or lower and the strongest around 6M HCl and higher. Using hydrochloric acid allows thorium and uranium to be separated from one another by using 6M HCl or higher to allow for uranium to stick to the resin and allowing thorium to pass straight through the column into waste. When using a nitric acid solvent at varying concentrations, the distribution coefficients for varying actinides are displayed in Table 1.4 [55]. Uranium absorbs very weakly to the resin at any concentration of HNO 3. Thorium absorbs the strongest between 6-10M HNO 3. Using nitric acid allows thorium and uranium to be separated by using ~8M HNO 3 to allow the thorium to stick to the column and let uranium to mostly pass through the column into waste. 19

33 Table 1.3: Distribution coefficients of actinides on anion exchange resins in pure HCl solutions. [55] 20

34 Table 1.4: Distribution coefficients of actinides on anion exchange resins in pure HNO 3 solutions. [55] Extraction chromatography uses the same basic ion exchange mechanism as ion exchange chromatography, but the stationary phase chemical makeup is more specific allowing for the chemical separations to be more selective. A list of extraction chromatography resin types and the actinides it can selectively separate are displayed in Table 1.5. UTEVA and TEVA resins will be focused on. The organic stationary phase for UTEVA is displayed in Table 1.6 and the log distribution coefficients plotted against HCl and HNO 3 concentrations for the UTEVA resin are shown in Figure 1.7 [56]. The organic stationary phase for TEVA is displayed in Table 1.6 and the log distribution coefficients plotted against HCl and HNO 3 concentrations for the TEVA resin is shown in Figure 1.8 [57]. The distribution coefficient charts for each stationary phase types can 21

35 be used to determine a separation protocol to be able to separate out the actinides of interests. Table 1.5: List of extraction chromatography resin types available commercially and the actinides it can selectively separate. Table 1.6: The organic stationary phase for UTEVA and TEVA extraction chromatography resin. [56-57] 22

36 Figure 1.7: UTEVA resin stationary phase log distribution coefficient for actinides plotted against the concentration of HNO 3 (left) and HCl (right). [56] 23

37 Figure 1.8: TEVA resin stationary phase log distribution coefficient for actinides plotted against the concentration of HNO 3 (left) and HCl (right). [57] 1.4 Research Focus The overall objective of this dissertation work was to determine what types of nuclear forensic signatures are important to analyze for in order to determine the origin of unknown samples. This was accomplished by analyzing a variety of uranium bearing materials (i.e. ore deposits, uranium metals, and soil samples from former metal fabrication facilities) in order to study unique nuclear forensic signatures, retrospectively relate unique signatures to historical chemical processing and origin of samples, evaluate sample preparation procedures for accuracy and precision, and introduce the relationship between trace elements and processes as potential signatures. In chapter 3, a nuclear forensic case study was developed around a particular uranium metal rod that was found 24

38 at an abandoned metal fabrication facility. Destructive and nondestructive analytical methods were used to analyze this uranium rod as well as soil samples that were collected from this facility. The main intention was to determine what signatures could be detected in both the metal and soil samples and see if they matched with the documented activity that once occurred at this site. In chapter 4, different etching procedures were investigated on a variety of uranium metal samples to determine if there would be any affect on their calculated radiological age. In chapter 5, soil samples from two former metal fabrication facilities were analyzed for signatures and determined if it was possible to distinguish the difference between the two facilities and then compared the signatures to the history of activity that occurred at each site. In chapter 6, physical or morphological characteristics, trace elemental make-up, and rare earth element analysis were performed on a variety of radiological materials (i.e. ore deposits, metals, soil). The data was used to determine unique characteristics in uranium bearing materials associated with the processes that occur throughout the various stages of the fuel cycle. These characteristics, when evaluated alone or in combination, become signatures. 1.5 References 1. Lawrence Livermore National Laboratory S&TR January/February Identifying the Source of Stolen Nuclear Materials, p Joint Working Group of the American Physical Society and the American Association for the Advancement of Science (2005) Nuclear forensics: role, state of the art, program needs. Washington, DC 3. Stanley FE, Stalcup AM, and Spitz HB (2013) J Radioanal Nucl Chem 295: Rye JS, Jeong YJ, Cha HJ, Shin HS, Cheong CS (2010) J Anal Sci Tech 1: Stefanka Z, Katona R, Varga Z (2008) J Anal At Spectrom 23:

39 6. Boulyga SF, Becker SJ (2001) Fresenius J Anal Chem 370: Boulyga, SF, Testa, C, Desideri, D, Becker JS (2001) J Anal At Spectrom 16: Bouyga, SF, Becker JS (2002) J Anal At Spectrom 17: Boulyga, SF, Matusevich JL, Mironov VP, et al (2002) J Anal At Spectrom 17: Meyers LM, Williams RW, Glover SE, LaMont SP, Stalcup AM, Spitz HB (2013) J Radioanal Nucl Chem 296: Boulyga SF, Prohaska T (2008) Anal Bioanal Chem 390: Srnik M, Steier P, Wallner G (2010) Nucl Instrum Methods Phys Res B 268: Carvalho FP, Oliveira JM (2010) Environ Int 36: Oliver IW, Graham MC, MacKenzie AB, Ellam RM, Farmer JG (2007) J Environ Monit 9: Browne E, Firestone RB, Shirley VS (1986) Table of Radioactive Isotopes. John Wiley & Sons, Inc., New York 16. Stirling CH (2012) Science 335: Hiess J, Condon DJ, McLean N, Noble SR (2012) Science 335: Zoriy MV, Kayser M, Izmer A, Pickhardt C, Becker JS (2005) Int J Mass Spectrom 242: Rich BL, Hinnefeld SL, Lagerquist CR, et al (1988) Health Physics Manual of Good Practices for Uranium Facilities, DOE 20. Moody KJ, Hutcheon ID, Grant PM (2005) Nuclear forensic Analysis Taylor & Francis, New York. 21. Mayer K, Wallenius M, and Ray I (2005) Analyst 130: Lamont SP, Hall G (2005) J Radioanal Nucl Chem 264: Williams RW, Gaffney AM (2011) Proc Radiochim Acta 1:

40 24. Vivone RJ, Godoy ML, Godoy JM, Santos, GM (2012) J Braz Chem Soc 23: Wallenius M, Morgenstern A, Apostolidis C, Mayer K (2002) Anal Bioanal Chem 374: Varga Z, Suranyi G (2007) Analytica Chimica Acta 599: Varga Z, Wallenius M, Mayer K, Hrnecek (2011) J Radioanal Nucl Chem 290: Moorthy AR, Kato WY (1995) Nuclear Materials Management 24: Rekha AK, Dingankar MV, Anilkumar S, Narayani K, Sharma DN (2006) J of Radioanal Nucl Chem 268: Taylor RN, Croudace IW, Warwick PE, Dee SJ (1998) Chem Geol 144: Boulyga SF, Becker JS, Matusevitch JL, Dietze, HJ (2000) Int J Mass Spectrom 203: Turner S, Calsteren PV, Vigier N, Thomas L (2001) J Anal At Spectrom 16: Shen CC, Edwards RL, Cheng H, et al (2002) Chem Geol 185: Shoji M, Hamajima Y, Takatsuka K, et al (2001) Appl Radiat Isotopes 55: Nir-El Y (2006) J Radioanal Nucl Chem 267: retrieved on Zhou W, Apkarian RP, Wang ZL, and Joy D. (2007) Scanning Microscopy for Nanotechnology, Springer, New York, Goldstein J, Newbury D, Joy D, et.al. (2003) Scanning Electron Microscopy and X- ray Microanalysis, 3 rd ed., Springer, New York 39. Scott JHJ (2003) Anal Bioanal Chem 375: retrieved on Albert B (2012) Methods in Physical Chemistry, 1 st ed., Wiley-WCH, Weinheim,

41 42. Behrens M and Schlogl R (2012) Characterization of Solid Materials and Heterogeneous Catalysts: From Structure to Surface Reactivity, 1 st ed., Wiley-WCH, Weinheim, Janssens K (2013) Modern Methods for Analyzing Archaeological and Historical Glass, 1 st ed., John Wiley & Sons, Supplies/Description/ retrieved on de Hoffmann E and Stroobant V (2007) Mass Spectrometry: Principles and Applications, 3 rd ed. Wiley, West Sussex 46. Skoog DA, Holler FJ, Crouch SR (2007) Principles of Instrumental Analysis, 6 th ed., Thomsom Brooks/Cole, Belmont, CA 47. Amr MA (2012) Adv Appl Sci Res 3: Douthitt CB (2008) Anal Bioanal Chem 390: Goldstein SJ and Stirling CH (2003) Rev in Mineral and Geochem 52: Wieser M, Schwieters J, and Douthitt C (2012) Isotopic Analysis: Fundamentals and Applications using ICP-MS, 1 st ed. Wiley-VCH, Weinheim, Wieser ME and Schwieters JB (2005) Int J Mass spectrum 242: Heumann KG, (2012) Isotopic Analysis: Fundamentals and Applications using ICP- MS, 1 st ed. Wiley-VCH, Weinheim, Grindler JE (1962) The Radiochemistry of Uranium, National Academy of Science/National Research Council 54. Hyde EK (1960) The Radiochemistry of Thorium, National Academy of Science/National Research Council 55. Korkisch J (1989) Handbook of ion exchange resins: Their application to inorganic analytical chemistry, Volume 2, CRC Press, Inc. 56. Horwitz EP, Dietz ML, and Diamond H (1992) Anal Chim Acta 266: Horwitz EP, Dietz ML, Chiarizia R, et al. (1995) Anal Chim Acta 310:

42 Chapter 2: Radiochronological Age of a Uranium Metal Sample from an Abandoned Facility 2.1 Introduction Nuclear forensics uses a variety of analytical methods and tools to explore the physical, chemical, elemental, and isotopic characteristics of nuclear and radiological material in order to determine its provenance, how it was manufactured, and other factors that may affect the material characteristics. The process of radiochronology typically employs several analytical methods that may include thermal ionization mass spectrometry (TIMS) [1-2], inductively coupled plasma mass spectrometry (ICP-MS) [3-8], gamma spectroscopy [9-11], and isotope dilution alpha spectrometry [12-13] to determine the isotopic composition and radiological age of a uranium-bearing material. The uranium decay series and the position of 230 Th and 234 U in the sequence of decay products is shown in Figure 2.1. The radiochronology (i.e., age dating) presented here determines the time since purification of a uranium-bearing material based upon the ingrowth of 230 Th toward its parent 234 U in a non-disturbed environment.. 29

43 Figure 2.1: 238 U decay chain series displays the decay of 234 U into 230 Th. In this work, a small sample of a piece of uranium metal rod buried for approximately fifty years in the dirt floor of an abandoned metal rolling milling was analyzed using multi-collector inductively coupled plasma mass spectroscopy (MC-ICP- MS). The uranium metal was likely a deformed piece or an end of a longer rod that was being rolled into a uniform diameter and length for eventual use in a nuclear reactor. A suitable sample for nuclear forensic analysis was obtained from the interior of the rod to minimize the possibility of exposure to environmental factors that could alter its physical and chemical properties. The 230 Th/ 234 U atomic ratio measured in this small sample of metal was used to determine the date when the uranium metal was last purified. Samples of the dirt in which the bar was buried were also analyzed and appear to be contaminated 30

44 with recycled uranium. The forensics analyses of the metal and the dirt lead to different conclusions on the origin of the uranium in these samples. This work was previously published in the Journal of Radioanalytical and Nuclear Chemistry [14]. 2.2 Experimental Materials All acids were from Seastar Chemical, Inc. (Sidney, BC, Canada). Poly prep chromatography columns and AG 1x8 resin bed ( mesh) were from Bio-Rad Laboratories, Inc. (Hercules, CA). TEVA resin was from EiChrom Technologies, LLC (Lisle, IL). The materials used for analysis in this work included a 233 U tracer solution calibrated with a natural uranium standard solution prepared from NBL CRM 112-A and a 229 Th tracer solution calibrated with the NIST SRM 4342A 230 Th radioactivity solution [3]. The TEVA resin was prepared by repeated suspensions in a centrifuge tube containing Milli-Q water, centrifugation, and removal of the foamy later using a transfer pipet. This process removes excess organic extraction reagent and produces a more uniform size distribution of the resin particles. The AG 1x8 ( mesh, chloride form) resin was prepared by repeated suspensions in Milli-Q water, allowing it to settle, and decanting and discarding any floating material. The resin was suspended twice in 6 M HCl, allowed to settle, and the acid was decanted and discarded. Finally, the resin was rinsed five times with Milli-Q water as described above Sample preparation of soil 31

45 The uranium isotopic content of three different samples of dirt were analyzed including: (1) the surrounding dirt in which the uranium metal rod was buried (raw soil), (2) the dust gently brushed from the surface of the rod (dust), and (3) yellow and white material scraped from the surface of the rod (scrapings). Samples of raw soil, dust, and scrapings were leached in nitric acid with addition of hydrofluoric acid, spiked with 233 U and 229 Th, and dried. The leachate containing uranium and thorium from each sample was purified and separated using a three-column procedure. First, anion-exchange chromatography was used to eliminate trace elements and organics (column A). Second, uranium was separated from the thorium using anion-exchange chromatography (column B). Third, the thorium fraction was purified using a TEVA resin bed (column C). The uranium fractions were re-dissolved in 3 ml of 2% HNO 3 for isotopic uranium analysis by MC-ICP-MS. The thorium fractions were re-dissolved in a 3 ml solution of 2% HNO M HF for isotopic thorium analysis by MC-ICP-MS Radiochemical separation and purification chromatography of soil samples Column A: Removal of Trace Elements Trace elements and organics were removed from the leachate samples using columns containing 1.8 ml of AG1x8 ( mesh) resin bed. The resin beds were pre-cleaned by rinsing with 10 ml of 0.1M HCl and 4 ml of water and then conditioned with 10 ml of 8M HNO 3. The samples (dissolved in 2 ml of 8M HNO 3 ) were transferred onto the columns and the sample vials were rinsed once with 2 ml of 8M HNO 3 and added to the columns. Iron was removed by rinsing the columns twice with 2 ml of 8M HNO 3. Uranium and thorium were eluted together into a 15 ml Teflon vial using 2 x 1 ml of 9M HCl, followed by 4 x 2 ml of a solution of 0.1M HCl M 32

46 HF and taken to dryness. Three drops of concentrated HNO 3 followed by two drops of concentrated HCl were added to each fraction and dried. Column B: Uranium and Thorium Separation Uranium and thorium were separated using columns containing a 1.0 ml of AG1x8 ( mesh) resin bed. The columns were conditioned by rinsing with 8 ml of 0.1M HCl and 6 ml of 9M HCl. A 15 ml telfon vial was placed under each column prior to sample loading to collect the thorium fractions. Samples from column A were dissolved in 0.5 ml of 9M HCl + 10 µl of concentrated HNO 3 and loaded onto the columns, along with a single 1 ml rinse of 9M HCl from each sample container. The thorium fractions were eluted with 5 ml of 9M HCl and dried. Three drops of concentrated HCl followed by two drops of concentrated HNO 3 was added to the thorium vials and dried. Uranium fractions were eluted into a 15 ml Teflon vial using 7 ml of 0.1M HCl and dried. Two drops of concentrated HNO 3 was added twice to the uranium vials and dried. Column C: Thorium Purification The final thorium purification step was performed using columns with 0.6 ml of TEVA resin. The columns were conditioned by rinsing with 2 ml of water and 4 ml of 4M HNO 3. Thorium samples from column B were dissolved in 0.5 ml of 4M HNO 3 and loaded onto the columns. Sample vials were rinsed with 0.5 ml and 1.0 ml of 4M HNO 3 and added to the columns. Residual uranium and other contaminants were removed by rinsing the columns twice with 2 ml of 4M HNO 3. Thorium was eluted with 1.5 ml of 9M HCl, and 8 ml of a solution of 0.1M HCl M HF solution into 15 ml Teflon vials. After drying the thorium factions, a series of drop additions were added to each 33

47 fraction and dried down: 1) two drops of concentrated HCl 2) three drops of concentrated HCl + one drop of concentrated HNO 3 and 3) two drops of concentrated HNO Sample preparation of uranium metal Surface oxidation on a small piece of the uranium metal rod taken from the interior was removed by soaking in a small volume of 8M HNO 3 followed by a series of rinses using 8M HNO 3, Milli-Q water, and then acetone. The dry, cleaned metal was weighed ( grams) and was dissolved in 4 ml of concentrated HNO 3 in a 30 ml Teflon vial on a hotplate until nitrogen dioxide vapors were no longer observed. A small amount of concentrated hydrofluoric acid was added to dissolve any residual precipitates. The sample was then diluted with 10 ml of 8M HNO 3 and 12 ml of Milli-Q water. A 250 µl sample of this primary solution ( grams) was further diluted to approximately 1 liter ( grams) with 2% HNO 3 and called the first dilution solution. This method follows that given in Williams and Gaffney [3]. For the 234 U fraction measurement, a weighed fraction of the first dilution was spiked with 233 U tracer, equilibrated by heating covered on the hotplate and then dried. This fraction was re-dissolved in 3 ml of 2% HNO 3 and was analyzed without further purification by MC-ICP-MS. For the 230 Th measurement, a weighed fraction of the primary dilution was spiked with a 229 Th tracer and subjected to four separation and purification steps (given below) using anion-exchange resin and TEVA resin prior to analysis. The separated and purified thorium fraction was re-dissolved in 3 ml of 2% HNO M HF solution and analyzed by MC-ICP-MS Radiochemical separation and purification chromatography of metal thorium fractions 34

48 Thorium was purified using standard techniques: initial purification was accomplished using column C (presented above); second, thorium was further purified from uranium using a 1.8 ml AG1x8 resin bed in 9 M HCl, on which uranium adsorbs and thorium passes (similar protocol to column B); third, thorium was absorbed on a 1 ml AG1x8 resin bed in 8 M HNO 3 and then eluted with 9 M HCl followed by 0.1 M HCl M HF (similar protocol to column A); and lastly, the thorium was passed through a 1 ml AG1x8 resin bed in 9 M HCl Instrumentation MC-ICP-MS The uranium and thorium mass spectrometric analyses were performed using a NuPlasma HR multi-collector ICPMS with a combination of Faraday and electron multiplier (pulse-counting) detectors. Samples were introduced to the plasma via a CETAC Aridus II system with a 100 ml/min Teflon nebulizer. Instrumental mass bias and detector cross-calibration factors for both the U and Th analyses were determined using a certified reference material (CRM U010) of 1% enriched uranium obtained from the U. S. Department of Energy New Brunswick Laboratory. The isotope dilution tracers, 233 U for uranium and 229 Th for thorium, were calibrated with gravimetrically prepared standard solutions of U metal (NBL CRM 112-A) and NIST SRM 4342A 230 Th radioactivity solution. Thorium was measured by peak-hopping 230 Th- 229 Th on an electron multiplier with simultaneous analysis of 232 Th on Faraday cups. Uranium was measured in static multi-collection mode with 234 U and 233 U on electron multipliers and 235 U and 238 U on Faradays. Memory effects were corrected by measuring blank acid solutions that were used to dissolve the purified samples immediately prior to sample 35

49 analysis. Data reduction involved correcting for detector baseline, acid blank, detector cross-calibration factor, and instrumental mass bias. 2.3 Results and Discussion The uranium isotopic composition, determined by MC-ICP-MS, for each of the samples analyzed is listed in Table 2.1 and is similar to that of natural uranium (i.e., 238 U = ± %; 235 U = ± %; 234 U = ± %) [15]. The lack of measureable 236 U in the uranium metal and the scraping samples indicates that these samples had not been subjected to nuclear fuel reprocessing or recycling. The presence of a small quantity of 236 U measured in the dust and raw soil samples may reflect contamination from recycled uranium associated with other work at the facility. Table 2.2 lists the total uranium mass and activity concentration determined for each sample. The piece of metal is 98.37% natural uranium and reflects a composition expected for uranium metal from this era. The yellow and white material scraped from the metal bar are likely uranium metal oxides that were formed on the surface of the metal bar while it was buried for over fifty years and exposed to ambient environmental conditions. 36

50 Table 2.1. Uranium isotopic composition in samples Sample Description 238 U, atom % 236 U, atom % A 235 U, atom % 234 U, atom % U Metal Scrapings ± 0.12 None detected ± ± U Metal Dust ± ± ± ± Raw Soil ± ± ± ± U Metal ± 0.14 None detected ± ± Natural Uranium B ± ± ± *U uncertainties are given as the combined standard uncertainty. A 236 U was determined to contribute less than 1E-6 atom percent based on the detection limit. B Corresponds to reference 15. Table 2.2. Total uranium concentration in samples Sample Description U, g/g µci/g U Metal Scrapings ± ± U Metal Dust ± ± Raw Soil ± ± U Metal ± ± *All uncertainties are the combined standard uncertainty. The ratio of 235 U/ 238 U determined for each sample is shown on Figure 2.2 where the solid line indicates the ratio for natural uranium ( ) and the dashed lines 37

51 indicate +/- 0.05% [16]. The low ratio observed for the raw soil may suggest that there is some contamination from depleted uranium, which had also been rolled at the facility. The uranium isotopic abundances found in the soil recovered from the dirt floor is very heterogeneous and most likely reflects the different types of metals that were rolled at the facility over time. Figure 2.2: Soil and uranium metal sample 235 U/ 238 U ratio comparison. The solid line indicates the ratio for natural uranium ( ) and the dashed lines indicate +/- 0.05% [15]. Error bars are the combined standard uncertainty. The mathematical expression for determining the time (t) since last purification, based upon the atom ratio of 230 Th to 234 U is: 1 t = (λ 234 λ 230 ) ln # 1+ N 230 Th % $ % N 234 U (λ 234 λ 230 ) λ 234 where λ 234, λ 230, N 234U, and N 230Th are the decay constants and number of atoms for 234 U and 230 Th, respectively. The number of atoms of uranium and thorium were determined & ( '( 38

52 from the isotope dilution MC-ICP-MS analysis. The half-lives used for 234 U and 230 Th are 245,250 ± 490 years (2σ) and 75,690 ± 230 years (2σ), respectively [17]. Figure 2.3: Age determination comparison of each sample. Error bars are expanded uncertainties (k=3). The calculated radiological age of each sample is listed on Figure 2.3 and has been determined from in-growth of 230 Th from radioactive decay of 234 U with the assumptions that (1) no 230 Th was present at the time the uranium metal was last purified, and (2) no thorium contamination was introduced after uranium purification. The relative age of uranium metal and the scraping are approximately the same (Fig. 3). The calculated age of the uranium metal is 61 years, corresponding to a purification date of July of 1950 ± 1.5 years. The calculated age of the scrapings is 60 years, corresponding to August of 1951 ± 3.5 years. The radiological age of the raw soil and dust is significantly less than the metal or scrapings and may be an artifact of the heterogeneity of the uranium contamination in the dirt floor, mobility of thorium and uranium due to 39

53 environmental transport, or housekeeping practices during plant operation that diluted or removed contamination in the floor. 2.4 Conclusions Radiochronology methods were used to analyze samples from an abandoned site in order to ascertain the material and isotopic composition and radiological age of suspected uranium-bearing materials. The uranium isotopic compositions measured in each sample is similar to that of natural uranium other than the slight 236 U detected in the dust and raw soil samples. The radiological age calculated for the raw soil and dust is significantly less than the samples of metal and scrapings. The difference in radiological age may be attributed to environmental transport or the results of clean-up activities to mitigate contamination while the facility was in operation. The findings from these analyses are evident that methods employed in radiochronology are valuable in ascertaining the provenance of intercepted materials. The age determination of the uranium metal was calculated to be 61 years corresponding to a production date of July of 1950 ± 1.5 years. The uranium metal is better suited for forensic analysis, as it is not as subject to the leaching and other factors which may fractionate thorium and uranium in soil samples. 2.5 References 1. Lamont SP, Hall G (2005) J Radioanal Nucl. Chem 264: Taylor RN, Croudace IW, Warwick PE, Dee SJ (1998) Chem Geol 144: Williams RW, Gaffney AM (2011) Proc Radiochim Acta 1:

54 4. Boulyga SF, Becker JS (2002) J Anal At Spectrom 17: Boulyga SF, Becker JS, Matusevitch JL, Dietze, HJ (2000) Int J Mass Spectrom 203: Boulyga SF, Becker JS (2001) Fresenius J Anal Chem 370: Turner S, Calsteren PV, Vigier N, Thomas L (2001) J Anal At Spectrom 16: Shen CC, Edwards RL, Cheng H, et al (2002) Chem Geol 185: Shoji M, Hamajima Y, Takatsuka K, et al (2001) Appl Radiat Isotopes 55: Nir-El Y (2006) J Radioanal Nucl Chem 267: Rekha AK, Dingankar MV, Anilkumar S, et al (2006) J Radioanal Nucl Chem 268: Wallenius M, Morgenstern A, Apostolidis C, Mayer K (2002) Anal Bioanal Chem 374: Boulyga SF, Testa C, Desideri D, Becker JS (2001) J Anal At Spectrom 16: Meyers LA, Williams RW, Glover SE, LaMont SP, Stalcup AM, Spitz HB. (2013) J Radioanal Nucl Chem 296: Browne E, Firestone RB, Shirley VS (1986) Table of Radioactive Isotopes. John Wiley & Sons, Inc., New York 16. Richter S, Alonso-Munoz A, Eykens R, et al (2008) Int J Mass Spectrom 269: Cheng H, Edwards RL, Hoff J, et al (2000) Chem Geol 169:

55 Chapter 3: Confidence in Radiological Chronometry of Uranium Metal Samples 3.1 Introduction Determining the age of nuclear or radiological material is an important signature in nuclear forensics because it reveals the elapsed time since the material was last purified. Radiological chronometry for uranium metals involves determining the uranium and thorium isotopic content of a sample and basing its age on the fractional ingrowth of the isotope, 230 Th, from the decay of its parent isotope, 234 U (Figure 2.1). Three chronometers can be used to determine the radiological age of uranium and thorium bearing material, viz., 230 Th/ 234 U [1-6], 228 Th/ 232 Th [7], and 231 Pa/ 235 U [8-9]. Use of the 230 Th/ 234 U chronometer assumes that all contaminants were removed when the material was last purified and requires complete separation of thorium from uranium and the removal of oxidation and residual contamination on the surface of the uranium metal sample prior to age determination. [7]. This research evaluated the impact of several etching procedures on the radiological age determined for multiple samples of uranium metal. The main objective of this work was to demonstrate the importance of rigorous etching of uranium metal samples to remove surface oxidation and other contamination that could introduce a bias to the determination of radiological age. The uranium metal rod, from which four small samples were removed for age determination, is shown in Figure 3.1. This rod was serendipitously found buried in the dirt floor of an abandoned metal rolling mill. Based upon the history of the mill, the rod was buried in the dirt floor for more than 50 years and was coincidently exposed to ambient, external environmental 42

56 conditions including rain, ice, and snow that could have altered the surface isotopic composition of uranium and thorium from that of the bulk metal. According to historical records, the uranium rod was likely last purified 63 years earlier with an uncertainty of only a few years [3]. Four small samples were cut from the rod (U-Rod) and were etched using three different procedures described in the experimental section. In addition to the uranium metal rod, small samples were removed from two different un-irradiated pure uranium metal nuclear fuel elements (F-element solid and F-element drillings), also displayed in Figure 3.1. These elements exhibited little, if any surface oxidation compared to the uranium metal rod. The etching procedures used with these samples are also described in the experimental section. Figure 3.1: Photographs of the uranium metal samples: Uranium rod (U-rod) (left), F- element Solid (middle), and F-element Drillings (right) 3.2 Experimental Materials All acids were from Seastar Chemical, Inc. (Sidney, BC, Canada). Poly prep chromatography columns and AG 1x8 resin bed ( mesh) were from Bio-Rad Laboratories, Inc. (Hercules, CA). TEVA resin was from EiChrom Technologies, LLC (Lisle, IL). The materials used for analysis in this work included a 233 U tracer solution 43

57 calibrated with a natural uranium standard solution prepared from NBL CRM 112-A and a 229 Th tracer solution calibrated with the NIST SRM 4342A 230 Th radioactivity solution. The TEVA resin was prepared by repeated suspensions in a centrifuge tube containing Milli-Q water, centrifugation, and removal of the foamy lather using a transfer pipet. The AG 1x8 ( mesh, chloride form) resin was prepared by repeated suspensions in Milli-Q water, allowing it to settle, and decanting and discarding any floating material. The resin was suspended twice in 6 M HCl, allowed to settle, and the acid was decanted and discarded Sample preparation of uranium metal The uranium and thorium isotopic content of three types of uranium metals were analyzed including: (1) four pieces of metal removed from the uranium metal rod obtained from the abandoned metal rolling mill (U-rod-1 through U-rod-4), (2) a small solid piece of metal cut from the end of the first un-irradiated fuel element (F-element solid), and (3) two very thin drilled turning aliquots from the second un-irradiated fuel element (F-element drillings 1 and F-element drillings 2). Surface oxidation on samples from the uranium metal rod was removed using three different etching procedures: Sample U-rod-1 was etched using aqua-regia followed by 8M HNO 3. Samples U-rod-2 and U-rod-3 were both separately etched using only 8M HNO 3. U-rod-4 was etched twice with 8M HNO 3 plus concentrated HCl. The metal sample from the first un-irradiated fuel element (F-element solid) was etched with 8M HNO 3. The metal drillings from the second un-irradiated fuel element (F-element drillings) were not etched for fear that they would completely dissolve in weak acid. After etching, samples were serially rinsed using 8M HNO 3, Milli-Q water, and acetone before dissolution. The dry, cleaned metal 44

58 samples were weighed and dissolved in concentrated HNO 3 with minimal HF to dissolve any residual precipitates. Each sample was diluted with Milli-Q water and 8M HNO 3 to obtain a 6M HNO 3 solution. A 250 µl aliquot of the bulk metal solutions were further diluted to approximately 1 liter with 2% HNO 3 and identified as primary dilution solutions following the method of Williams and Gaffney [2]. If necessary, secondary dilution solutions were prepared by taking a 500 µl aliquot of the primary dilution metal solutions and further diluting to approximately 10.5 ml with 2% HNO 3. To determine the 234 U and uranium concentrations, a weighed fraction of the secondary dilution solution of each metal sample was spiked with 233 U as a tracer, equilibrated by heating covered on the hotplate, and then dried. These fractions were redissolved in 3 ml of 2% HNO 3 and analyzed without further purification by multicollector inductively coupled plasma mass spectrometry (MC-ICP-MS). Uranium isotopic ratio analyses were performed using 20 µl of each primary dilution solution dissolved in 3 ml of 2% HNO 3 and analyzing without further purification by MC-ICP- MS. Measurements of 230 Th were made by taking a weighed fraction of each primary dilution, adding 229 Th tracer, and performing three separation and purification steps (described below) using anion-exchange resin and TEVA resin prior to analysis. The separated and purified thorium fractions were re-dissolved in 3 ml of 2% HNO M HF solution and analyzed by MC-ICP-MS. The samples could not be analyzed in triplicate due to constraints in resources Radiochemical separation and purification chromatography of thorium metal fractions Column A: Thorium Separation 45

59 Thorium was separated from the sample matrix using chromatography columns containing 1.8 ml bed of AG1x8 ( mesh) resin. The columns were conditioned by rinsing with 6 ml of 9M HCl. A 15 ml Teflon vial was placed under each column prior to sample loading to collect the thorium fraction. Samples, dissolved in 3 ml of 9M HCl + 25 µl of concentrated HNO 3, were loaded onto the columns, along with two 2 ml rinses of 9M HCl of each sample container. The thorium fractions were eluted with an additional 4 ml of 9M HCl and dried. Three drops of concentrated HNO 3 were added twice to the Teflon vials and dried. Column B: Thorium Purification The first thorium purification step was performed using chromatography columns loaded with 0.6 ml of TEVA resin. The columns were conditioned by rinsing with 4 ml of 4M HNO 3. Thorium fractions from column A were dissolved in 1.0 ml of 4M HNO 3 and loaded onto columns containing TREVA resin.. Residual thorium in the vials was added to the columns by rinsing the vials first with 0.5 ml and then with 1.0 ml of 4M HNO 3. Residual uranium and other contaminants were removed from the chromatography column by rinsing twice with 2 ml of 4M HNO 3. Thorium was eluted from the columns into 15 ml Teflon vials using 1.5 ml of 9M HCl and 6 ml of a solution consisting of 0.1M HCl M HF. Three drops of concentrated HCl were added to the thorium in the Teflon vials and dried followed by three drops of concentrated HNO 3 being added twice and dried. Column C: Final Thorium Purification The final purification of the thorium fraction from column B was performed using columns containing 1.0 ml of AG1x8 ( mesh) resin. The columns were 46

60 conditioned by rinsing with 6 ml of 9M HCl. A 15 ml Teflon vial was placed under each column prior to sample loading to collect the thorium fractions. Samples were dissolved in 1.0 ml of 9M HCl + 25 µl of concentrated HNO 3 and loaded onto the columns, along with a 1 ml and 2 ml rinse of 9M HCl from each sample container. The thorium fractions were eluted with an additional 2 ml of 9M HCl and dried. Three drops of concentrated HNO 3 were added twice to the thorium vials and dried Instrumentation MC-ICP-MS The uranium and thorium mass spectrometric analyses were performed using a NuPlasma HR multi-collector ICPMS with a combination of Faraday and electron multiplier (pulse-counting) detectors. The details associated with the instrumentation procedures and specifications were previously reported [3]. 3.3 Results and Discussion The reliability of radiological age determination using the fractional ingrowth of a decay product from a pure parent radionuclide is very dependent upon the accuracy and precision of the uranium and thorium isotopic measurement results, the purity and homogeneity of the material being analyzed, the presence of contamination from environmental or industrial sources, and whether surface oxidation and contaminants have been thoroughly removed from the samples prior to analysis. Figure 3.2 (and tabulated in Table 3.1) gives the results of age determination using the 230 Th/ 234 U chronometer for samples of uranium metal that were etched using various procedures, outlined above in the experimental section. U-rod-1 was subjected to the most rigorous etching procedure [3] and exhibited the most reliable radiological age based upon the 47

61 history of the mill. Less rigorous etching used with U-rod-2 and U-rod-3 yielded radiological ages much older than the known age. U-rod-4, which was subjected to concentrated HCl as an etching agent, appeared much older than the others. The use of hydrochloric acid is thought to remove mainly uranium oxidation complexes from the metal surface leaving behind the thorium oxidation complexes; this leads to a radiological age appearing much older since the 234 U/ 230 Th ratio would not be reflective of the bulk metal. During the uranium metal casting operation, it has been reported in the literature that a significant portion of thorium becomes separated from the uranium matrix and enriched towards the surface as the molten solidifies, allowing for heterogeneity to occur throughout the uranium metal [10]. Figure 3.2: Age determination comparison of each sample. Error bars are expanded uncertainties (k=3). 48

62 Table 3.1: Predicted model age of each uranium metal sample Sample description U, g/g Model Production Date U-rod ± July 06, 1950 ± 1.59 years U-rod ± Dec. 09, 1938 ± 1.67 years U-rod ± Apr. 10, 1936 ± 1.71 years U-rod ± Nov. 08, 1908 ± 2.33 years F-element Solid ± May 21, 1933 ± 1.81 years F-element Drillings ± years old ± 8.12 years F-element Drillings ± years old ± 2.96 years * U uncertainties are given as the combined standard uncertainty. * Reference date for all metal samples is March 19, 2012 besides U-rod-1: Aug. 1, 2011, and U-rod-4: Aug. 7, 2012 The technical basis of radiological chronometry requires that the material be very pure when it was last processed since the radiological age is determined from the fractional ingrowth of a relatively short-lived radioactive decay product from a pure longlived parent material. The 234 U/ 230 Th chronometer assumes that all the thorium was removed when the uranium metal was last purified. Residual thorium contamination will bias the 234 U/ 230 Th ratio and will yield an older age. [1]. Table 3.2 illustrates that the thorium content of U-rod-1 is much less than the other U-rod samples, thus demonstrating that its age accurately reflects the true age of the material compared to the other U-rod samples. In fact, U-rod-4 shows the highest thorium concentration of all the U-rod samples and its age appears very biased. 49

63 Table 3.2: Thorium isotopic and concentration for each uranium metal sample The radiological ages of the un-irradiated uranium metal nuclear fuel elements were older than expected, a consequence of reduced effectiveness of the etching procedures than that used for U-rod-1. The unetched F-element drillings 1 and 2 exhibited a radiological age hundreds of years older than the expected age, similar to the U-rod-4 sample. Thus, rigorous etching with concentrated HNO 3 may be necessary for removing thorium and other contaminants from the oxidized surfaces. Radiological chronometry was used by McCulloch to determine the age of a piece of uranium metal, certified reference material CRM112A (also identified as NBL960) to use as a standard to compared against other geological samples. The age of CRM112A determined by McCulloch was ~36 years older than the actual age of the standard reference material [11]. A similar finding was reported for this standard by Lawrence Livermore National Laboratory and displayed in Figure 3.3. Currently, CRM112A is only certified for uranium concentration and uranium isotopic composition. Although it can be used for radiological chronometry, its age is not certified. The date provided on 50

64 the certificate represents when the uranium content was approved and gives no indication when the uranium metal was purified or whether thorium was removed. The certificate gives a procedure for etching the uranium metal prior to analysis using 8M HNO 3 [12]. Results obtained in this research shows that 8M HNO 3 alone is not sufficient to remove surface oxidation and will yield a biased estimate of the radiological age. More samples of CRM112A should be analyzed using the rigorous etching described in this research to estimate the radiological age of the standard uranium metal material. If the uranium CRM112A metal was sufficiently purified to remove residual thorium, the estimated radiological age determined after rigorous etching should more accurately reflect the duration of time since the material was last purified. Figure 3.3: McCulloch and LLNL age dating of CRM112-A comparison. 3.4 Conclusions 51

65 Radiological chronometers used for age determination of uranium metal samples are one of several important signatures for nuclear forensics. Rigorous etching with concentrated HNO 3 appears necessary to remove surface uranium and thorium that may differ in isotopic composition from the bulk material. This research demonstrates that surface etching protocols need to be further evaluated to avoid introducing a systematic bias in determining radiological age. Radiological ages could appear older than expected in the presence of excess 230 Th obtained at time of fabrication or incomplete separation of uranium and thorium at time of analysis. The availability of an uranium metal standard without thorium contamination and a rigorous etching procedure could potentially lead to the development and availability of age dating reference materials that are currently unavailable for nuclear forensics. 3.5 References 1. Lamont SP, Hall G (2005) J Radioanal Nucl Chem 264: Williams RW, Gaffney AM (2011) Proc Radiochim Acta 1: Meyers LA, Williams RW, Glover SE, LaMont SP, Stalcup AM, Spitz HB. (2013) J Radioanal Nucl Chem 296: Vivone RJ, Godoy ML, Godoy JM, Santos, GM (2012) J Braz Chem Soc 23: Wallenius M, Morgenstern A, Apostolidis C, Mayer K (2002) Anal Bioanal Chem 374: Varga Z, Suranyi G (2007) Analytica Chimica Acta 599: Varga Z, Wallenius M, Mayer K, Hrnecek (2011) J Radioanal Nucl Chem 290: Moorthy AR, Kato WY (1995) Nuclear Materials Management 24: Rekha AK, Dingankar MV, Anilkumar S, Narayani K, Sharma DN (2006) J of Radioanal Nucl Chem 268:

66 10. DOE Ohio Field Office Recycled Uranium Project Final Report, May 15, McCulloch MT, Mortimer GE (2008) Australian Journal of Earth Sciences 55: New Brunswick Laboratory, DOE, Certificate of Analysis CRM 112-A, Sept. 30,

67 Chapter 4: Uranium Isotopic Signature Analysis of Collected Dirt Samples at Former Metal Rolling Mills 4.1 Introduction Uranium isotopic composition is an important nuclear forensic signature that can reveal information about the history of the material, how it was used, or the type of processes that were involved in fabricating the material. There are three naturally occurring uranium isotopes with long half-lives, viz., 238 U, 235 U, and 234 U. The isotopic composition of natural uranium is ± % 238 U, ± % 235 U, and ± % 234 U by mass [1]. Whether a sample of uranium is natural, depleted, or enriched in the 235 U isotope can be determined by evaluating the 235 U/ 238 U mass ratio. The most common value for the natural 235 U/ 238 U mass ratio is [2-4]. Uranium used in nuclear fuel and weapons generally requires enrichment of the 235 U isotope significantly above that in natural uranium. The presence of 236 U in a sample is another important isotopic signature that suggests the uranium may have been reprocessed after exposure in a nuclear reactor [5]. This research investigated the nuclear forensics signatures in samples of dirt collected at two different facilities in the United States at which uranium metal was processed as part of the historical atomic weapons program. The uranium isotopic content and concentrations of uranium and plutonium were determined in samples of soil from both facilities. The objective of the research was to investigate the relationship between these measured signatures and the known processes and activities that formerly were conducted at these sites and establish a technical basis of information that could be 54

68 used to interpret activities at other concealed or otherwise restricted facilities. Retrospective analysis of the uranium isotopic composition in these samples provides a comprehensive set of data to relate isotopic signatures to processes and activities associated with the production and use of nuclear or radiological materials. 4.2 Experimental Materials All materials not listed in this section were previously listed in section U- TEVA resin was acquired from EiChrom Technologies, LLC (Lisle, IL). A 242 Pu tracer solution calibrated with an in-house LLNL Pu-239 standard solution prepared from Pumetal, and the calibration was confirmed with NIST SRM 4334H Pu Methods The concentration and isotopic composition of uranium, thorium, and plutonium were determined in samples of dirt by first acid leaching aliquots of the dirt and then separating the chemical fractions using ion exchange chromatography. The chemical yield of each fraction was determined using isotopic tracers. Multi-collector inductively coupled plasma mass spectrometry (MC-ICP-MS) was used to determine the isotopic composition of separated fractions. The samples were collected from two very different facilities. The first facility was a metal rolling mill at which hot forged, mostly natural uranium metal bars were rolled between the 1940 s and 1950 s into specific dimensional sizes for use in the early nuclear reactors. This facility was abandoned around the 1980 s and has physically 55

69 deteriorated such that rain, snow and other debris have leaked through the roof and broken windows and mixed with the dirt floor of the mill. A contemporary sample of the dirt floor near one of the rolling mills in this facility was collected and analyzed in this research. The other site from which dirt was collected was the former Fernald Feed Material Plant where uranium metal ingots were fabricated from various uranium feed materials. Samples were collected from the burn area around the former Fernald fire station as well as material from the K-65 waste silo. Samples analyzed in this research were assigned the following identification labels: SSA: Mill Dirt #1 SSB: Mill Dirt #2 FSA: Fernald Soil #1 FSB: Fernald Soil #2 F Waste: Fernald K-65 Silo Waste Sample Preparation for U/Th Analysis The uranium and thorium isotopic signatures in all samples were analyzed by first dissolving the dirt, soil, or waste in nitric acid (with addition of hydrofluoric acid) and hydrochloric acid/perchloric acid followed by the addition of 233 U and 229 Th isotopic tracers. The dissolved solutions were then dried and redissolved in 3 ml of 9M HCl + 25 µl HNO 3. Uranium and thorium were separated and purified using a four-column procedure. First, anion-exchange chromatography was used to separate uranium from thorium (column A). Second, the uranium fraction from column A was further purified using a U-TEVA resin bed (column B). Third, the thorium fraction from column A was purified using a TEVA resin bed (column C). Lastly, the thorium fraction from column C was further purified using anion-exchange resin bed (column D). The uranium fractions were re-dissolved in 3 ml of 2% HNO 3 for isotopic uranium analysis by MC- 56

70 ICP-MS. The thorium fractions were re-dissolved in a 3 ml solution of 2% HNO M HF for isotopic thorium analysis by MC-ICP-MS. Soil samples were not analyzed in triplicate because of time and resource limitations U/Th Radiochemical separation and purification chromatography Column A: Uranium from Thorium Separation Uranium and thorium were separated using columns containing 2.2 ml of AG1x8 ( mesh) resin bed. The columns were conditioned by rinsing with 6 ml of 9M HCl. A 15 ml Teflon vial was placed under each column prior to sample loading to collect the thorium fractions. Samples, dissolved in 3 ml of 9M HCl + 25 µl of concentrated HNO 3, were loaded onto the columns, along with two 2 ml rinses of 9M HCl from each sample container. The thorium fractions were eluted with an additional 4 ml of 9M HCl and dried. Three drops of concentrated HNO 3 were added twice to the thorium vials and dried. Uranium fractions were eluted into a 15 ml Teflon vial using 10 ml of 0.1M HCl and dried. Three drops of concentrated HNO 3 were added to the uranium vials and dried. Column B: Uranium Purification Uranium fractions from column A were purified using columns containing 1.0 ml of U-TEVA resin bed. The resin beds were conditioned with 3 ml of 4M HNO 3. Samples, dissolved in 0.5 ml of 4M HNO 3, were loaded onto the columns, along with a 0.5 ml and 1 ml rinse of 4M HNO 3 from each sample container. Residual contaminants were removed by rinsing the columns with an additional 2 ml of 4M HNO 3, 1.5 ml of 9M HCl, followed by 4 ml of 5M HCl. Uranium fractions were eluted into a 15 ml 57

71 Teflon vial using 6.5 ml of 0.1M HCl and dried. Three drops of concentrated HNO 3 were added twice to the uranium vials and dried. Column C: Thorium Purification The first thorium purification step was performed using columns with 0.6 ml of TEVA resin. The columns were conditioned by rinsing with 4 ml of 4M HNO 3. Thorium fractions from column A were dissolved in 1.0 ml of 4M HNO 3 and loaded onto the columns. Sample vials were rinsed with 0.5 ml and 1.0 ml of 4M HNO 3 and added to the columns. Residual uranium and other contaminants were removed by rinsing the columns twice with 2 ml of 4M HNO 3. Thorium was eluted with 1.5 ml of 9M HCl and 6 ml of 0.1M HCl M HF solution into 15 ml Teflon vials. Three drops of concentrated HCl were added to the thorium vials and dried. Three drops of concentrated HNO 3 were added twice to the vials and dried. Column D: Final Thorium Purification Final purification of the thorium fraction from column C was performed using columns containing 1.0 ml of AG1x8 ( mesh) resin bed. The columns were conditioned by rinsing with 6 ml of 9M HCl. A 15 ml Teflon vial was placed under each column prior to sample loading to collect the thorium fractions. Samples were dissolved in 1.0 ml of 9M HCl + 25 µl of concentrated HNO 3 and loaded onto the columns, along with a 1 ml and 2 ml rinse of 9M HCl from each sample container. The thorium fractions were eluted with an additional 2 ml of 9M HCl and dried. Three drops of concentrated HNO 3 were added twice to the thorium vials and dried Sample Preparation for Pu Analysis 58

72 Additional samples of the dirt, soil, and waste were digested using a protocol similar to the protocol for U/Th analysis. A 244 Pu isotopic tracer was added to the digested samples. The dried samples were dissolved in 5 ml of 8M HNO 3 and were separated and purified using a five-column procedure, described below. The separated and purified plutonium fractions were re-dissolved in a 3 ml solution of 2% HNO M HF for isotopic plutonium analysis by MC-ICP-MS. Samples were not analyzed in triplicate because of time and resource limitations Pu Radiochemical separation and purification chromatography Column E: Plutonium Purification 1 An initial plutonium separation and purification step was performed with the plutonium fractions using columns containing 2.0 ml of AG1x8 ( mesh) resin bed. The columns were conditioned by rinsing with 10 ml of 8M HNO 3. Samples were loaded onto the columns along with two 2 ml rinses of each sample container using 8M HNO 3. The columns were converted to chloride form by rinsing with 8 ml of 9M HCl. A 15 ml Teflon vial was placed under each column to collect the plutonium fractions. The plutonium fractions were eluted with double 2 ml rinses of a 9M HCl + HI solution (15 parts of 9M HCl to 1 part concentrated HI) and twice with a 4 ml rinse of 9M HCl + HI solution, and dried. Each plutonium fraction received a series of additions of (1) two drops of concentrated HCl, (2) three drops of concentrated HCl + one drop of concentrated HNO 3 and 3) two drops of concentrated HNO 3. Samples were dried between each addition. Column F: Plutonium Purification 2 59

73 The second purification step was performed using columns with 1.0 ml of TEVA resin. The columns were conditioned by rinsing with 4 ml of 4M HNO 3. Plutonium fractions from column E were dissolved in 1.0 ml of 4M HNO µl of NaNO 2 and loaded onto the columns. Sample vials were rinsed twice with 1.0 ml of 4M HNO 3 and added to the columns. Other contaminants were removed by rinsing the columns twice with 2 ml of 4M HNO 3. Plutonium was eluted with 1 ml of 9M HCl, 3 ml of 0.1M HCl M HF solution, and 6 ml of 0.1M HCl + HI solution (12 parts of 0.1M HCl to 1 part concentrated HI) into 15 ml Teflon vials. Three drops of concentrated HCl were added to the plutonium vials and dried. Three drops of concentrated HNO 3 were added to the vials and dried. Column G: Plutonium Purification 3 The third purification step was performed using columns containing 0.3 ml of TEVA resin. The columns were conditioned by rinsing with 2 ml of 4M HNO 3. Plutonium fractions from column F were dissolved in 0.5 ml of 4M HNO µl of NaNO 2 and loaded onto the columns. Sample vials were rinsed with 0.25 ml and 0.5 ml of 4M HNO 3 and added to the columns. Other contaminants were removed by rinsing the columns twice with 1 ml of 4M HNO 3. Plutonium was eluted with 0.75 ml of 0.1M HCl M HF solution and 2.5 ml of 0.1M HCl + HI solution (12 parts of 0.1M HCl to 1 part concentrated HI) into 15 ml Teflon vials. Three drops of concentrated HCl were added to the vials and dried. Thereafter, three drops of concentrated HNO 3 were added to the vials and dried. Column H: Plutonium Purification 4 60

74 The dried plutonium fractions from column G were purified again using the procedure for column F. Column I: Final Plutonium Purification A final purification step of the dried plutonium fractions from column H was performed using columns containing 1.0 ml of AG1x8 ( mesh) resin bed. The columns were conditioned by rinsing with 8 ml of 0.1M HCl and 6 ml of 9M HCl. Samples were dissolved in 0.5 ml of 9M HCl + 10 µl concentrated HNO 3 and loaded onto the columns plus a 1 ml rinse of 9M HCl from each sample container. Uranium and other contaminants were removed by rinsing the columns with 1 ml and 2 ml of 9M HCl. A 15 ml Teflon vial was placed under each column to collect the plutonium fractions. The plutonium fractions were eluted with a total of 7 ml of a 9M HCl + HI solution (15 parts of 9M HCl to 1 part concentrated HI). Three drops of concentrated HCl were added to the plutonium vials. Two drops of concentrated HNO 3 were then added to the dried sample and dried again Instrumentation MC-ICP-MS The uranium and thorium mass spectrometric analyses were performed using a NuPlasma HR multi-collector ICPMS with a combination of Faraday and electron multiplier (pulse-counting) detectors. The details associated with the instrumentation procedures and specifications were previously reported [7]. 4.3 Results Uranium isotopic analyses of soil from each of these facilities provide unique signatures to distinguish the operations conducted at these two uranium facilities. The 61

75 235 U/ 238 U ratio (Figure 4.1) for the dirt and soil samples demonstrates that soil from Fernald has a slightly enriched uranium signature while the dirt from the floor of the rolling mill represents a signature more consistent with natural uranium. The solid line in Figure 4.1 is the conventional value for natural uranium and is shown for reference. Soil samples from Fernald exhibit a unique 236 U signature that was not found at the rolling mill (Figure 4.2). The total uranium and plutonium concentrations, listed on Table 4.1, suggest that the facilities had very different functions in fabrication of uranium for the early nuclear reactors. While the rolling mill samples have higher uranium content, the samples of soil collected at Fernald have a higher plutonium concentration. Figure 4.1: 235 U/ 238 U ratio comparison for soil samples (Solid line indicates ratio for natural uranium) 62

76 Figure 4.2: 236 U/ 235 U vs. 238 U/ 235 U ratio comparison plot for soil samples Table 4.1. Uranium and plutonium concentrations in soil samples. Sample description U, g/g Pu ( 239 Pu+ 240 Pu), pg/g Pu/U ratio (ppb) SSA ± ± ~ SSB ± ± ~ SSC ± Not Analyzed NA FSA ± ± 1.5 ~67 FSB ± ± 1.5 ~76 F Waste ± ± ~ * U uncertainties represent the combined standard uncertainty. The Fernald waste, which is a combination of residual materials from many different processes involving pitchblende ores and wastes from several uranium processes, exhibits a unique signature of 234 U content (Figure 4.3) being much greater than that found in the Fernald soil or dirt from the rolling mill. The concentration of total uranium in the Fernald waste is similar to that of the Fernald soil, but lacks any detectable 236 U. The Fernald waste also contains minimal detectable plutonium. 63

77 Figure 4.3: 234 U/ 235 U vs. 238 U/ 235 U ratio comparison plot for soil samples 4.4 Discussion While the concentration of uranium in dirt from the rolling mill was much greater than that measured in the Fernald soil, the isotopic signatures of enriched uranium and the presence of 236 U in Fernald soil clearly suggest that processes conducted at Fernald and the rolling mill were different. The 236 U signature indicates that some of the uranium feed material processed at Fernald was recycled and had, at some previous time, been in a reactor. A small quantity of 236 U was also detected in one of the samples from the mill suggesting that at some time during its history a batch of recycled uranium was rolled at the mill. The higher plutonium content in the samples of Fernald soil is further evidence that some of the uranium processed at the facility had been recycled. Plutonium found in recycled uranium is produced by neutron capture in 238 U [8-9]. Thus, a small amount of plutonium may remain in the recycled material [10]. Uranium was reclaimed at Fernald from all the waste. These raffinates were burned to remove organics, which effectively concentrates plutonium in the residues. In 1985, a DOE Task Force evaluated the 64

78 processing of recycled uranium at several facilities and reported that Fernald processed recycled uranium containing more plutonium than expected (i.e., more than 10 parts of plutonium per billion parts of uranium on a mass basis) [11]. The Pu/U ratio measured in the Fernald soil (Table 4.1) is a unique signature for this facility and confirms recycled uranium supplied to the plant as feed material contained excess plutonium. According to the Fernald Plant history, natural, depleted, and some enriched uranium was fabricated into ingots for fuel and target rods. Within the plant, uranium scraps, residues, slag, chips, and cleaning solvents from the refining processes were recycled through the process to avoid loss of product. Although analyses of soil exhibited signatures associated with low-enriched and recycled uranium, the signature for depleted uranium was not observed. Only a limited number of samples of Fernald soil were analyzed for this research so the results are likely not representative of all processes conducted at the site. The sample obtained from the Fernald K-65 waste silo, which contained a mixture of many different types of process waste, exhibited a concentration of 234 U greater than the soil. The slight enrichment of 234 U is a characteristic of the cascade in which it was produced [12]. The abandoned mill rolled mostly hot forged natural uranium metal into dimensional rods. However, the history of the mill also shows that a small quantity of depleted and slightly enriched uranium was also rolled before the facility was closed. Soil from the mill exhibited a signature most closely related to natural uranium. There was little if any plutonium detected in the samples. 4.5 Conclusions 65

79 Nuclear forensics uses a variety of analytical methods and tools to evaluate the physical, chemical, elemental, and isotopic characteristics of nuclear and radiological material. These characteristics, when evaluated alone or in combination, become signatures that may allow determination of the materials provenance and the method of manufacture. This research demonstrates that the isotopic composition of uranium and plutonium measured in samples of soil collected at a facility suspected or known to be processing uranium can reveal important details about the actual work being performed at the site. The ratio of 235 U/ 238 U measured in soil at two different uranium processing facilities was found to indicate whether natural, depleted, or enriched uranium was processed. Furthermore, the presence of 236 U and isotopes of plutonium in the soil was found to indicate whether recycled or reprocessed uranium material was being utilized at the plant. This work has demonstrated that there is a relationship that connects the measured signatures to the known history of processes and activities that were formerly conducted at these types of facilities. This type of information can be used as a technical basis of that could be used to interpret activities at other concealed or otherwise restricted facilities. 4.6 References 1. Browne E, Firestone RB, Shirley VS (1986) Table of Radioactive Isotopes. John Wiley & Sons, Inc., New York 2. Stirling CH (2012) Science 335: Hiess J, Condon DJ, McLean N, Noble SR (2012) Science 335: Zoriy MV, Kayser M, Izmer A, Pickhardt C, Becker JS (2005) Int J Mass Spectrom 242: Boulyga SF, Prohaska T (2008) Anal Bioanal Chem 390:

80 6. Williams RW, Gaffney AM (2011) Proc Radiochim Acta 1: Meyers LM, Williams RW, Glover SE, LaMont SP, Stalcup AM, Spitz HB (2013) J Radioanal Nucl Chem 296: Tandon L, Kuhn K, Martinez P, et al (2009) J Radioanal Nucl Chem 282: Wallenius M, Lutzenkirchen K, Mayer K, et al (2007) J Alloys Compd : Rich BL, Hinnefeld SL, Lagerquist CR, et al (1988) Health Physics Manual of Good Practices for Uranium Facilities, DOE 11. S. Cohen & Associates (Feb. 2011) SC&A Review of Issues Related to Reconstruction of Doses for Workers Exposed to Recycled Uranium at Fernald A Second White Paper 12. Moody KJ, Hutcheon ID, Grant PM (2005) Nuclear forensic Analysis Taylor & Francis, New York. 67

81 Chapter 5: Elemental and Auxiliary Nuclear Forensic Signatures in a Variety of Uranium Bearing Materials 5.1 Introduction Analyzing for trace impurities in uranium bearing materials from different stages throughout the fuel cycle is vital for nuclear forensics. Trace impurities can be added intentionally to nuclear material for process reasons or to improve certain properties of the final product. Impurities can also be introduced into the material unintentionally as contaminants from the uranium feed material or from the processes used throughout the fuel cycle to convert uranium ore into targeted items. As material is processed farther in the fuel cycle, there are fewer contaminants introduced inadvertently since they are gradually removed in the different production stages [1]. The presence of trace elements at the part per million (ppm) concentration levels in nuclear fuel plays an important part of the fuel s properties and performance. Strong neutron absorbers such as Cd, B, Gd, Sm, Dy, and Eu can affect the functional properties of nuclear fuel by absorbing neutrons causing the fuel to be less efficient at producing energy in nuclear power reactors. Elements such as Al, Ca, Cr, Fe, Ni, Mn, Pb, and Si can affect the metallurgical properties of the fuel. For this reason, quantification of trace impurities in nuclear grade uranium material is a routine quality control release requirement for use in nuclear reactors [2]. Probing for rare earth elements (REE) in radiological materials associated with the fuel cycle is also essential for forensic purposes; it may provide details about the origins of materials or provide insight into production processes. REE, also known as 68

82 lanthanides, consists of Y, La, Ce, Pr, Nd, Pm, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, and Lu. This group of elements is naturally bundled together in nature because of their similar trivalent oxidation states and ionic radii. Exceptions to this trend include Ce and Eu, which can be in the +4 or +2 oxidation state respectively under some environmental conditions. REE are very abundant in the earth s crust and tend to act comparable in the environment with their similar chemical properties; this makes separating one REE from another very difficult [3]. In the determination of individual lathanides in samples, it is important to note that the natural abundances of individual lanthanides follow the Oddo-Harkin s rule, which states the abundance of elements with even atomic number is greater than elements with odd atomic number. Hence, a plot of the REE concentration verse increasing atomic number (Z) displays a zigzag rather than a smooth line [4, 5]. To correct for this issue, REE abundances are typically normalized to C1 chondrites, which are stony meteorites, as they are thought to be similar in composition to the earth s mantle [3, 6]. An example of what the REE patterns looks with and without normalization can be seen in Figure

83 Figure 5.1. Rare earth element concentrations plotted against Z number with and without normalization. Displayed is actual data presented later on in this chapter. A variety of uranium bearing materials such as uranium metals, uranium ore deposits, and soil samples from former metal fabrication facilities were analyzed for unique characteristics that could potentially be used as nuclear forensic signatures. Physical characteristics (e.g. morphology of samples), trace elements, and REE analyses were examined for each of the samples. 5.2 Experimental Materials All acids were from Seastar Chemical, Inc. (Sidney, BC, Canada). Poly prep chromatography columns and AG 1x8 resin bed ( mesh) were from Bio-Rad Laboratories, Inc. (Hercules, CA). TEVA resin was from EiChrom Technologies, LLC (Lisle, IL). The materials used for analysis in this work included a 233 U tracer solution calibrated with a natural uranium standard solution prepared from NBL CRM 112-A and a 229 Th tracer solution calibrated with the NIST SRM 4342A 230 Th radioactivity solution. The TEVA and UTEVA resin was prepared by repeated suspensions in a centrifuge tube containing Milli-Q water, centrifugation, and removal of the foamy lather using a transfer pipet. The AG 1x8 ( mesh, chloride form) resin was prepared by repeated suspensions in Milli-Q water, allowing it to settle, and decanting and discarding any floating material. The resin was suspended twice in 6 M HCl, allowed to settle, and the acid was decanted and discarded Surface Analysis Techniques 70

84 SEM/EDS Scanning electron micrographs (SEM) were collected by a FEI InspectF-D8886 coupled with energy-dispersive X-ray spectroscopy (EDX) working with an accelerating voltage of 20 kv. Uranium metal samples were placed on a metal stub containing carbon tape and positioned onto a sample holder for SEM/EDS analysis. Samples were conductive and no coating was necessary. Ore samples were placed on a metal stub, coated with epoxy, and polished down to make a flat surface for analysis. Samples were coated with around 8nm of iridium to make samples conductive and positioned onto a sample holder for SEM/EDS analysis XRD Uranium ore samples were ground down using a mortar and pestle and ~200 mg of sample was placed on the XRD sample holder with small additions of acetone to keep particles in place. Samples were air dried in preparation for XRD analysis. The powder ore samples were analyzed on a Bruker AXS D8 ADVANCE X-ray diffractometer equipped with a LynxEye 1-dimentional linear Si strip detector. DIFFRACplus Evaluation package Release 2009 software was used for the data analysis. The unknown samples were scanned from θ. The step scan parameters were 0.02 step and 2 second counting time per step with a 15mm variable divergence slit and a 1.0 anti-scatter slit. The samples were x-rayed with Ni-filter Cu radiation from a sealed tube operated at 40kV and 40mA. Phases in the unknown samples were identified by comparison of observed peaks to those in the International Centre for Diffraction Data (ICDD PDF2009) powder diffraction database. X-ray reference material (Bruker supplied 71

85 Al 2 O 3 ) was analyzed during the time of the unknown runs to ensure goniometer alignment. No peak shift was observed in the reference material Sample Preparation Methods Uranium and thorium sample preparation for ore samples The uranium and thorium isotopic content of two types of ore samples were analyzed including: (1) Joachimsthal pitchblende ore, and (2) western US ore. Ore samples were dissolved using 8M and concentrated nitric acid (with a small addition of hydrofluoric acid) and hydrochloric acid/perchloric acid drying treatments. Final bulk solutions were brought up in 4M HNO 3. For uranium isotopic analysis, a weighed fraction of the bulk solution for each ore sample was spiked with 233 U tracer, equilibrated by heating covered on a hotplate, and dried. The dried aliquots, were dissolved in 1 ml of 9M HCl + 25 µl concentrated HNO 3 and were separated and purified using a two-column procedure. The uranium aliquots were first put through a modified 1.8 ml anion-exchange resin bed (column A, presented in section 5.2.4) that separated uranium from thorium; the uranium portion was collected while the thorium went to waste. Second, the uranium fraction from column A was further purified using a U-TEVA resin bed (column B, presented in section 5.2.4). The uranium fractions were dried and re-dissolved in 3 ml of 2% HNO 3 for isotopic uranium analysis by MC-ICP-MS. For thorium isotopic analysis, a weighed fraction of the bulk solution for each ore sample was spiked with 229 Th tracer, equilibrated by heating covered on a hotplate and 72

86 then dried. The dried aliquots, were dissolved in 1 ml of 9M HCl + 25 µl concentrated HNO 3 and were separated and purified using a three-column procedure. The thorium aliquots were purified using a 1.8 ml AG1x8 resin bed in 9 M HCl (modified column A, presented in section 4.2.4), in which uranium adsorbs onto the column and thorium passes through; second, thorium went through a TEVA resin bed (column C, as presented in section 5.2.4); and lastly, the dried thorium fraction was passed through a 1 ml AG1x8 resin bed (column D, as presented in section 4.2.4). The thorium fractions were dried and re-dissolved in a 3 ml solution of 2% HNO M HF for isotopic thorium analysis by MC-ICP-MS Uranium metal and soil sample preparation The uranium metal samples were prepared according to Chapter 3, section 3.2. The soil samples were prepared according to Chapter 4, section Elemental and rare earth elemental analysis sample preparation To acquire elemental composition data on ore, soil, and uranium metal samples, aliquots from each bulk sample solutions were transferred to a 15 ml Teflon vial, dried, and diluted with 2% HNO 3. To obtain the rare earth element (REE) composition of ore, uranium metal, and F-waste samples, aliquots from each bulk sample solution were transferred to a 15 ml Teflon vial, dried, and put through a 1.8 ml anion exchange column to perform a group separation in which the uranium matrix stays on the column resin and the rare earth elements elutes off the column immediately. Columns containing 1.8 ml of AG1x8 ( mesh) resin were conditioned by rinsing with 6 ml of 9M HCl. A 15 ml Teflon vial was placed under each column prior to sample loading to 73

87 collect the REE fractions. Samples, dissolved in 1 ml of 9M HCl + 25 µl of concentrated HNO 3, were loaded onto the columns, along with 1 ml and 2 ml rinses of 9M HCl from each sample container. The REE fractions were eluted with an additional 4 ml of 9M HCl and dried. Three drops of concentrated HNO 3 were added twice to the REE vials and dried. The dried REE fractions were re-dissolved in ~10 ml of 2% HNO 3 in preparation for ICP-MS analysis. A REE standard was prepared using aliquots of rare earth element standard reference solutions and normalizing each element to the C1 chondrite normalization factors for each rare earth element according to literature values by Anders and Grevesse [7]. This particular set of normalization factors were used because they were reported to provide the smoothest REE pattern for a variety of geological samples compared to other published values [6]. The C1 chondrite standard was analyzed by ICP-MS as a quality control sample to account for sample recovery, correct concentration values for each element, and contamination issues that may arise. After ICP-MS analysis, all sample results were normalized to the using C1 chondrite normalization factors [7] to account for the odd/even abundance effects for this group of elements Destructive Instrumentation Techniques MC-ICP-MS The uranium and thorium mass spectrometric analyses were performed using a NuPlasma HR multi-collector ICP-MS with a combination of Faraday and electron multiplier (pulse-counting) detectors. The details associated with the instrumentation procedures and specifications were previously presented in Chapter 2, section

88 Quad-ICP-MS A fully quantitative REE and elemental analysis using both non-matrix and matrix-matched (for the metal samples) linear calibration curves based on NIST traceable standards was performed using a Thermo Electron XSeries Quadrupole Inductively Coupled Plasma Mass Spectrometer. The internal standard corrects for instrument drift and suppression from the matrix. Interference corrections were applied to spectral interferences, both isobaric and polyatomic. Elements such as As and Se were run in CCT mode (Collision Cell Technology) to remove polyatomic interferences. 5.3 Results and Discussion Uranium metal surface comparison The physical morphology and trace elemental composition for radiological materials are two important categories for nuclear forensic signature analysis. In Figure 5.2, a scanning electron microscopy (SEM) image of the oxidized side of U-rod-1 is shown to the left and the shiny metallic surface side of the same metal piece is shown to the right. The dark regions of the oxidized side reveal the presence of trace impurities in the sample (determined using EDS). On the metallic surface SEM image, small crystals, like the one circled, can be seen towards the middle of the image that was formed from the heating and cooling cycles associated with the uranium metal fabrication process. The presence of crystalline impurities may be key signatures that can be used along with other evidence to track the unknown location and fabrication processes used to produce uranium metal materials. 75

89 Figure 5.2. SEM images of U-rod-1 oxidized side (left) demonstrates the presence of trace impurities and the shiny metallic side shows crystal formations indicated by black circle (right). The physical characteristics of five other metal samples using SEM are presented in Figure 5.3. The oxidized surfaces of U-rod-1 (A) and U-rod-2 (B) have very closely related characteristics in which there are heterogeneous dark and light colored sections within the metal pieces, indicating the presence of impurities. U-rod-3 (C) has similar characteristics with dark and light colored spots as well as contrasting stripes. The F- element solid sample (D) has very smooth features and appears to contain very little impurities shown by a comparable contrast of gray shaded regions throughout the sample. The F-element drillings 1 (E) and 2 (F) samples look alike, in which linear structured columns of ordered material are present. There are some dark colored impurity sections in both samples that can easily be seen in the images. Electron dispersive spectroscopy (EDS), was used to determine qualitative elemental composition data for each metal sample. The dark impurity spot located in Figure 5.2 is made up of nickel. The darker colored impurity sections in the F-element drillings 1 and 2 samples are made up of aluminum, most likely from the cladding material in which the uranium was encased. 76

90 Figure 5.3. SEM images of uranium metal samples for comparison: A. U-rod-1, B. U- rod-2, C. U-rod-3, D. F-element Solid, E. F-element Drillings 1, F. F-element Drillings Uranium metal elemental analysis A list of the most abundant trace elements found in each dissolved sample by ICP-MS is tabulated in Table 5.1. Similar trace elements were also detected on the metal surface indicated by EDS. The most abundant trace elements in the samples are Na, Zn, Fe, Mg, Al, Ca, and Ni in no particular quantitative order or sample. According to literature, the impurities in pure uranium metals that were processed in the 1940s and 1950s include Fe through Ni noted two lines above in addition to Si, Mn, and Cr [8-10]. Ca, Mg, and Na were commonly used with halides during the production processing of uranium metal [11] so their presence in the samples are to be expected. While these uranium metal samples vary in trace element abundances, they can be used to distinguish one uranium metal type from the other and may help indicate where the uranium metals 77

91 were each rolled or fabricated. The location of where each uranium metal was rolled or fabricated is unknown. In general, the F-element Drillings 1 and 2 seem to have the largest abundance of impurities compared to the other uranium metal samples indicated by the EDS prescreening. This is most likely due to the lack on sample etching that took place during the sample preparation, previously presented in Chapter 3, or may also represent impurities from the metal cladding. Elemental composition comparisons for each metal sample that were presented in Table 5.1 are displayed by using radar plots shown in Figure 5.4. U-rod-2, U-rod-3, and F-element Solid samples have similar plots in which there is a large abundance of Fe, Ni, Ca, and K. F-element Drillings 1 and 2 samples are most abundant in Zn. The F-element Drilling 1 sample is also abundant in Al, which is most likely from the aluminum cladding that encased the metal sample. To further distinguish between the uranium metal samples, uranium isotopic signatures were also analyzed and compared. All uranium metal samples are comprised of natural uranium (Table 5.2). The F-element shavings 1 and 2 samples differ from the other metal samples by having a higher 236 U and 234 U content demonstrated by the correlation plots in Figure 5.5 and Figure 5.6, respectively. These two uranium metal samples are fabricated with some recycled uranium material indicated by the high 236 U content. The slight enrichment of 234 U content is most likely a characteristic from the cascade used during an enrichment stage in which the material was processed by [11]. 78

92 Figure 5.4. Elemental radar plots for element comparison in uranium metal samples. The numbers indicate normalized elemental concentrations (µg/g ) for each sample. 79

93 Figure 5.5. Uranium isotopic ratio correlation plot demonstrating F-element Drillings 1 and 2 have significant 236 U signature. Figure 5.6. Uranium isotopic ratio correlation plot demonstrating F-element Drillings 1 and 2 have significant 234 U signature. 80

94 Table 5.1. Elemental composition of uranium metal samples by ICP-MS. 81

95 Table 5.1 continued: 82

96 Table 5.2. Uranium isotopic composition and concentration in uranium metal samples. 83

97 5.3.3 Ore sample surface analysis Analyzing uranium ore samples for uranium isotopic information, elemental composition, and morphology characteristics are important when determining the provenance of ore deposits and other uranium bearing materials. They may have distinct signatures that can relate current and/or future uranium bearing materials to the initial uranium source origin. Two different types of uranium ore deposits from different geolocations were analyzed: (1) Joachimsthal pitchblende ore and (2) western US ore. Examining the physical appearances, phases, and elemental composition of the ore samples can help identify more dissimilarity features. The western ore sample looks more like a heterogeneous rock with an uneven yellow surface and the pitchblende ore has a shiny black homogenous surface seen by the naked eye. When probing the physical appearances in SEM images displayed in Figure 5.7, the western ore sample has a rough and ridged surface (left) while the pitchblende ore sample has a smooth surface (right). The phases of each ore sample were analyzed by XRD. The XRD spectrum for the western US ore, presented in Figure 5.8, displays the major phases including: quartz (SiO 2 ), fluorite (CaF 2 ), and iriginite ([U(MnO 4 ) 2 (OH) 2 ] 2H 2 O). The spectrum for the pitchblende ore sample gave a large broad peak for the pitchblende ore, shown in Figure 5.9. When utilizing EDS to gather qualitative elemental composition information, there were an abundant number of elements, including most of the rare earth elements and other metals such as Ca, Fe, and Ti. The broad peaks in the spectrum are most likely attributable to interference patterns between several phases within the ore deposit or have an amorphous phase structure [13]. 84

98 Figure 5.7. SEM images of Western ore (left) and Pitchblende ore (right). Figure 5.8. XRD spectrum of western US ore sample. 85

99 Figure 5.9. XRD spectrum of pitchblende ore sample Uranium Ore Elemental Analysis Differences in uranium isotopic and uranium concentration can help distinguish one ore type from the other. The uranium characteristics in both ore samples are natural indicated by the 235 U/ 238 U ratio presented in Table 5.3. The western US ore has a higher 234 U content (Figure 5.10) and the pitchblende ore has a larger uranium concentration, also displayed in Table

100 Figure Uranium isotopic ratio correlation plot indication Western US ore has a higher 234 U signature. Table 5.3. Uranium concentration comparison between ore samples. Ores U, g/g 238 U/ 235 U Ratio Western Ore ± E-03 ± 3.84E-06 Pitchblende Ore ± E-03 ± 3.85E-06 There are a wide variety of trace elements in various ore deposits that can be linked to certain geological regions in the world [14]. The major trace elements found in the western US ore and the pitchblende ore are listed in Table 5.4. To make it easier to compare the elemental compositions for each ore sample, the data from Table 5.4 was manipulated and plotted into radar plots, shown in Figure The Joachimsthal pitchblende ore sample that was analyzed was rich in calcium, which corresponds well with calcite and dolomite minerals. It is also abundant in Fe and Ti. According to 87

101 literature [15], pitchblende ore in general, is commonly associated with Fe, Cu, Co, Pb, Ag, and Bi as trace elements. In the Joachimsthal region, the ore deposits tend to be pitchblende and rich in Si and Co. Minerals such as quartz, calcite, and dolomite are also associated to this region. The particular geo-location of the western US ore that was analyzed is unknown, though the sample is abundant in Al, K, Ca, and Fe. In the Colorado Plateau, the ore deposits tend to be yellow in color, consists of carnotite phases, and have copper-uranium type deposits as well. Similar deposits are also located near Garo, Park County, Colorado and near San Rafael Swell area in Utah. In Utah, the ore deposits tend to be plentiful in Fe, Cu, Co, U, F, and V minerals. Ore deposits located near Grants District in New Mexico are mostly comprised of uranophane and tyuyamunite phases (which consist of calcium complexes), as well as hematite and limonite phases (which are consist of iron complexes). The origin of the western ore sample that was analyzed would seem to be from New Mexico due to the ore deposits being rich in Ca and Fe in this area. Figure Elemental radar plots for element comparison in uranium ore samples. The numbers indicate normalized elemental concentrations (mg/g) for each sample. 88

102 Table 5.4. Elemental composition of uranium ore samples by ICP-MS. 89

103 Table 5.5. Elemental composition of former metal fabricating soil samples by ICP-MS. 90

104 5.3.5 Soil, Uranium Metal, and Ore Sample Elemental Comparison Soil samples from Fernald and the abandoned rolling mill were also analyzed for elemental composition. The most abundant elemental compositions for the soil samples are displayed in Table 5.5. The data from Table 5.5 was plotted into radar plots, shown in Figure 5.12, to make it easier to compare the elemental composition between the samples. Soil in general, is made up of organic and inorganic minerals and contains various types of metals. The main elements found in the soil samples include Na, Mg, Al, K, Ca, Ti, and Fe. (Note: the organic elements such as C, N, O, P, and S were not analyzed along with Si.) Samples SSA and SSB plots are almost identical and are most abundant in Fe and Ca. Similar results are found with FSA and FSB samples, which are rich in Mg, Al, K, Ca, and Fe. These elements are most likely a combination of contaminants from the processes that occurred at each facility as well as major elements that can be found in natural soil. Natural soil in the US mostly contains these sample elements as well as Mn, Cu, and Zn [16, 17]. Pb, Ti, and Na were the major elements found in the F-waste sample but not in the other samples. This was most likely a side product from the reprocessing waste that was collected from the Manhattan project along with Fernald s own reprocessing waste. 91

105 Figure Elemental radar plots for element comparison in former metal rolling facilities soil samples. The numbers indicate normalized elemental concentrations (mg/g ) for each sample Figure 5.13 displays a venn diagram of the most abundant elements, determined by ICP-MS, that overlap with all three types of samples including soil, uranium metals, and uranium ores that have been presented in this chapter so far. Al, K, Ca, Pb, Fe, and Ti are all the elements that overlap between the uranium metals, ores, and soil samples. There are many questions that still remained unanswered relating to how/why these elements relate to one another in each of the sample types. More samples need to be analyzed that are connected linearly between ore deposits to uranium targeted items in which the ore deposit s geo-location is known, more soil samples from different metal fabrication facilities, and a deeper understanding of the literature and nuclear fuel cycle 92

106 processes. This chapter has demonstrated that it may possible to link major elements between the different sources along the fuel cycle but identifying those elements is the next step. This will be discussed in more detail in chapter 6. Figure Venn Diagram of common elemental make-up between sample types Rare Earth Elemental Analysis Rare earth elemental analysis was accomplished on uranium metal, ore, and F waste samples. Identifying specific REE patterns for different types of ore deposits from different geographical locations may be beneficial markers or signatures for nuclear forensic purposes. A group separation (previously discussed in section ) was done using anion exchange chromatography allowing the REEs to be separated from the 93