Article pubs.acs.org/ac
Mass Spectrometric Determination of Uranium and Thorium in High Radiopurity Polymers Using Ultra Low Background Electroformed Copper Crucibles for Dry Ashing Isaac J. Arnquist,* Eric J. Hoppe, Mary Bliss, and Jay W. Grate* Pacific Northwest National Laboratory, P.O. Box 999, Richland, Washington 99352, United States S Supporting Information *
ABSTRACT: A rapid new method for determining the U and Th mass concentrations in high radiopurity plastics is described, consisting of (1) dry ashing the plastic sample and tracers in low mass crucibles made of ultra low background electroformed copper (ULB EF-Cu) foil cut and folded into boats, (2) dissolving both the ash and the boat in acid, (3) performing a column separation to remove copper, and (4) determining the elements of interest by isotope dilution mass spectrometry. This method was demonstrated on both unfluorinated and fluorinated plastics, demonstrating high tracer recoveries and detection limits to pg/g (i.e., parts per trillion) levels or below, corresponding to μBq/kg of material. Samples of biomedical polyester (Max-Prene 955) and a fluoropolymer (polyvinylidene fluoride, PVDF) were analyzed in powder raw material forms as well as solids in the form of pellets or injection molded parts. The polyester powder contained 6 pg/g and 2 pg/g for 232Th and 238U, respectively. These levels correspond to 25 and 25 μBq/kg radioactivity, respectively. Determinations on samples of PVDF powder were typically below 1 pg/g for 232Th and 2 pg/g for 238U, corresponding to 4 and 25 μBq/kg radioactivity, respectively. The use of low mass ULB EF-Cu boats for dry ashing successfully overcame the problem of crucible-generated contaminants in the analysis; absolute detection limits, calculated as 3 × standard deviation of the process blanks, were typically 20−100 fg within a sample set. Complete dissolution of the ash and low mass boat provided high tracer recoveries and provides a convincing method to recover both the tracer and sample isotopes when full equilibration of tracer isotopes with sample isotopes is not possible prior to beginning chemical sample processing on solids.
I
The radiopurities of plastic materials from past campaigns for ULB physics detectors range from many mBq/kg to hundreds of μBq/kg for typical materials. Selected examples from the literature are given in Table SI_1 in the Supporting Information, with results reported in radioactivity units and/ or mass units. (The radioactivity units in this table are mBq/kg, in contrast to analytical results in the remainder of the paper that will be presented in μBq/kg). Measurement methods included gamma spectroscopy, nuclear activation analysis (NAA), and mass spectrometry. Also shown for some examples are the masses and times involved in gamma counting determinations, typically involving kg of material and weeks of counting time. Even so, the radiopurity levels from gamma spectroscopy are often reported as “less than” values in the mBq/kg range; gamma spectroscopy is not practical for determining radiopurities in the low μBq/kg range. For reference, 1 pg/g 232Th and 238U are equivalent to 4.1 and 12.4 μBq/kg, respectively.8,10 The mass and time requirements for determinations by gamma spectroscopy seriously limit material screening. In
nterest in the determination of metal elements in solid plastics has intensified in recent years. The European Union, in an effort to restrict hazardous substances in materials and waste electronic equipment, has established limits for heavy metals in plastics and recycled materials. In response, certified reference materials (CRMs) have been developed for metals such as Pb, Sb, Cd, Cr, and Hg in plastics such as polyethylene, polypropylene, polyvinyl chloride, and polyester.1−5 Concern over heavy metals in plastics is motivated by human health and the environment. For quite different reasons, the purities of plastic materials are a concern in the ultralow background (ULB) physics community.6−15 In constructing detectors designed to look for rare events, such as dark matter interactions or neutrinoless double beta decay, it is imperative to use materials of high radiopurity so that radionuclides such as 238U and 232Th and their daughter products6,7 do not elevate detector background levels. In terms of mass, μg/g to ng/g (ppm to ppb) levels of metals are important for consumer plastics, whereas mitigating 232Th and 238U to ng/g to pg/g or below (ppb to ppt or below) is important for ULB detector applications in rare event physics. Thus, requirements for radiopure plastics for detector applications far exceed the requirements for chemical purity in industrial or consumer settings. © XXXX American Chemical Society
Received: December 6, 2016 Accepted: February 14, 2017 Published: February 14, 2017 A
DOI: 10.1021/acs.analchem.6b04854 Anal. Chem. XXXX, XXX, XXX−XXX
Article
Analytical Chemistry addition, to support next generation ULB physics experiments, more materials of different types must be screened to find candidates with more desirable radiopurities and to ensure that materials batches and fabrication approaches maintain radiopurity within background budgets. In addition to being timeconsuming, gamma spectroscopy no longer provides data at sufficiently low detection limits. As a result, mass spectrometry, specifically inductively coupled plasma mass spectrometry (ICPMS), has emerged as a method that can provide extremely low detection limits from small masses in short periods of time.6,8,10,16−18 ICPMS has more than sufficient sensitivity to determine ultra low levels, provided (1) one can get the metal atoms out of the solid polymer matrix and into the mass spectrometer as ions using a process of sample preparation and source introduction, (2) the sample preparation method does not contaminate the sample and raise backgrounds, and (3) a suitable quantification method is available. In principle, metal determination could consist simply of dry ashing the polymer, dissolving the ash, and performing ICPMS with liquid nebulization.6,10,19,20 However, this preparation must be done without biasing the results with introduced contaminants. A recent panel discussion on ICPMS emphasized the limitations of contaminants in noncleanroom environments and from digestion vessel materials, and noted, “... the extremely low detection limits of ICP-MS have pushed the purity of available reagents and consumables, ... Plastics, glass, and even disposable sample preparation materials have to be f ree of trace metal contamination”.21 Pt crucibles, for example, are generally known for poor purity, and analyses by Nisi et al. confirm this.10 Fused silica or quartz crucibles can be degraded by ashing polymers, either by chemical etching or devitrification; these structural changes may release contaminants from the container material into the sample.17,22 Thus, sample preparation, with low process blank levels, remains a major challenge to assaying detector materials to meet stringent radiopurity goals. A further challenge for the quantification of metals in plastics has been the lack of standard reference materials for external calibration. The emergence of plastic CRMs has enabled progress in the analysis of heavy metals in plastics by laser ablation (LA) ICPMS.3,4,23 Nevertheless, there are no CRMs for U or Th in plastics and certainly not down to the pg/g (i.e., parts per trillion) levels of interest. For determinations of U and Th in detector materials, internal calibration using non-natural or nonabundant isotopes, i.e., 233U and 229Th tracers, have been adopted to get a ratiometric determination of the U and Th isotopes of interest.16,17 This method is typically referred to as isotope dilution mass spectrometry (IDMS), although to be rigorous, the tracer isotopes and sample isotopes must be equilibrated prior to further sample preparation steps, a requirement which cannot always be met. Isotope dilution is generally regarded as the most accurate and precise quantitation method for ICPMS analyses and is formally referred to as a primary ratio method of measurement and sometimes as a definitive method.24−30 This ratiometric method compensates for chemical effects such as sample preparation losses and instrumental drift, under the assumption that the two isotopes of an element behave identically through the process and measurements (i.e., without fractionation). “In general it can be stated that IDMS is the most important reference method for elemental analysis, offering highest accuracy and precision or smallest measurement uncertainties, when properly applied.”28
Our interests are in plastics for structural applications in ULB detectors where an insulating material is required, and in analytical methods with sufficiently low detection limits to assay the ULB polymers of interest. It is imperative that the method determine the sample U and Th concentrations, rather than assaying contaminants derived from crucible materials. Here we describe a new technique where ULB electroformed (EF) copper (Cu) (i.e., ULB EF-Cu) in thin foil form is cut and folded into boats, shown in Figure 1, to serve as very low mass
Figure 1. (a) Electroformed Cu foil and Cu boats and (b) ICPMS assay method for U and Th in plastic, using boats of ULB EF-Cu for the dry ashing step.
crucibles for dry ashing. Electroforming is the method preferred for obtaining the highest radiopurity copper as a material for UBL detector construction, with purity levels below 8.4 and 10.6 fg/g (ppq) for 232Th and 238U, respectively, in bulk samples.16 (For these applications, electroforming is performed in underground laboratories to limit cosmogenic formation of radionuclides from the copper, a process we can perform at PNNL.) The overall sequence of sample preparation in this new method includes adding tracers to mg quantities of polymer, dry ashing in a tube furnace, dissolution of the ash and the Cu boat material together, and a column-based separation to remove copper prior to liquid nebulization into an ICPMS, using clean room operations.(Figure 1) The low mass ULB EFCu foil boats, completely dissolved as part of the sample preparation sequence, may be regarded as perhaps the ultimate highly pure disposable sample preparation containers. They can be produced in large quantities in whatever shape or size the analyst requires from sheets of ULB EF-Cu foil, and a large number of samples can be ashed in parallel. The Supporting Information includes Figure SI_1 showing images of the overall process of preparing the ULB EF-Cu foil boats and carrying out the assay. B
DOI: 10.1021/acs.analchem.6b04854 Anal. Chem. XXXX, XXX, XXX−XXX
Article
Analytical Chemistry
Acquisition times were typically 10−30 s for each isotope of interest (i.e., 229Th, 232Th, 233U, and 238U). Three signal acquisitions were performed for each sample. Relative standard deviations for 232Th and 238U of ∼5% were typical in this study. A set of typically three process blanks with tracer were run at the same time with each set of samples, where a process blank consists of an ULB EF-Cu boat that goes through the exact same procedure at the same time as the samples. If studying the addition time of tracer on the samples, process blank sets were set up with tracer spiking before and after ashing. Process blanks were averaged to determine the background to subtract from sample data, and their standard deviation was used in the determination of absolute detection limits, calculated as 3 × standard deviation of the process blanks. Absolute detection limits of 20−100 fg are typical within a sample set. Concentration-based relative detection limits were then calculated based on the mass of polymer sample used. Polymer samples determined below their respective relative detection limit were assigned the maximum contaminant concentration values as an upper limit, represented with a preceding “
