Journal of Radioanalytical and Nuclear Chemistry

, Volume 296, Issue 3, pp 1303–1310

Surface cleaning techniques: ultra-trace ICP-MS sample preparation and assay of HDPE

Authors

    • Pacific Northwest National Laboratory
  • E. W. Hoppe
    • Pacific Northwest National Laboratory
  • R. S. Addleman
    • Pacific Northwest National Laboratory
Article

DOI: 10.1007/s10967-012-2301-1

Cite this article as:
Overman, N.R., Hoppe, E.W. & Addleman, R.S. J Radioanal Nucl Chem (2013) 296: 1303. doi:10.1007/s10967-012-2301-1

Abstract

The world’s most sensitive radiation detection and assay systems depend upon ultra-low-background (ULB) materials to reduce unwanted radiological backgrounds. In this study, we evaluate methods to clean HDPE, a material of interest to ULB systems and the means to provide rapid assay of surface and bulk contamination. ULB-level material and ultra-trace-level detection of actinide elements is difficult to attain, due to the introduction of contamination from sample preparation equipment such as pipette tips, sample vials, forceps, etc and airborne particulate. To date, literature available on the cleaning of such polymeric materials and equipment for ULB applications and ultra-trace analyses is limited. For these reasons, a study has been performed to identify an effective way to remove surface contamination from polymers in an effort to provide improved instrumental detection limits. Inductively Coupled Plasma Mass Spectroscopy was utilized to assess the effectiveness of a variety of leachate solutions for removal of inorganic uranium and thorium surface contamination from polymers, specifically high density polyethylene (HDPE). Leaching procedures for HDPE were tested to optimize contaminant removal of thorium and uranium. Calibration curves for thorium and uranium ranged from 15 ppq (fg/mL) to 1 ppt (pg/mL). Detection limits were calculated at 6 ppq for uranium and 7 ppq for thorium. Results showed the most effective leaching reagent to be clean 6 M nitric acid for 72 h exposures. Contamination levels for uranium and thorium found in the leachate solutions were significant for ultra-low-level radiation detection applications.

Keywords

HDPELeachingUltra-traceICP-MSUraniumThorium

Introduction

Ultra-low-background (ULB) radiation detection systems have achieved remarkable sensitivity in recent years. This capability is important for basic science experiments aimed at detecting rare events, such as next-generation neutrinoless double-beta decay and dark matter experiments [19].

Ultra-low-background systems and their performance then depend upon developing methods to minimize and reduce radiological contamination. With the incorporation of cleaner starting materials, radiological background signals are lowered, improving signal-to-noise ratios and ultimately lowering detection limits. Cleaning and passivation of a variety of detector materials for the removal of radiological contaminants in support of ULB detection capabilities has been shown to be a key area of research and development [4, 1014]. Applications such as high-purity germanium radiation detectors frequently rely upon polymers as a support material while scintillation detectors use polymeric materials as the detector itself. In some cases polymers would also be advantageous to use for shielding, containment of detection media or cryogenic fluids [4]. Because polymeric materials are typically composed of elements that have no natural long lived radioisotopes they are ideal ULB materials, unfortunately contaminants typically render their activities too high for utilization.

In order to efficiently and permanently remove elevated backgrounds present in such polymeric materials for ULB applications two critical elements must first be identified, the origin of the contamination (to avoid further introduction) and a successful removal process to produce materials of enhanced radiopurity. Contaminants may be introduced into the polymeric matrix via a number of different routes. Dust deposition perhaps enhanced by electrostatic attraction can transport fine particulates with U and Th. Liquids used in processing can deposit fine particulate and soluble radionuclides of Ra and U. The majority of the reported polymeric materials of interest are being investigated due to their inherently low radioactive contamination [4, 14]. In this sense, the ability to use polymeric materials for ULB systems is then limited only by the ability to remove and measure the contaminants (i.e. 238U, 232Th, and relevant progeny) in the polymeric materials of interest as well as performing forensic investigations into the source(s) of contamination to prevent future exposures. Assay of ultra-trace levels of inorganic contaminants in an organic polymer matrix is a non-trivial challenge.

Preparation (and supporting assay) methods to clean surface contamination from HDPE and assay via ultra-trace level Inductively Coupled Plasma Mass Spectroscopy (ICP-MS) have been evaluated. For this type of analysis, ICP-MS is an ideal assay method. It provides a high degree of elemental and isotopic specificity, a large dynamic range, and trace-level detection limits. Further, ICP-MS assays are more rapidly accomplished for long-lived radionuclides when compared to nuclear counting methods since prolonged counting times are avoided without a significant loss of sensitivity. However, even as such, ICP-MS measurements especially in the areas of trace level radionuclide assay are plagued with background contamination from naturally occurring radionuclides [12, 1517], making pre-cleaning of labware critical for successful analyses. Additionally, the presence of transition metal elements and oxides introduced during sample preparation can also cause a variety of interferences to be present [18, 19], which if eradicated could significantly lower detection limits. ICP-MS techniques such as incorporation of a collision cell provide the means to greatly reduce these interferences via breakdown of polyatomic ions [20]. However, in the ultra-trace-level detection regime where instrument sensitivity is of high importance, the addition of a reaction gas into the collision cell serves to further dilute the sample, and thus the signal reaching the detector. When analyte signal is sufficiently low (<200 ppq), long dwell times (>5 s/mass) [21] are required to achieve low standard deviations and reduce noise [22]. In cases such as this, diluting the sample through use of a reaction cell is undesirable. The analytical focus then shifts from instrumental modification to sample preparation techniques.

To date, research on ULB polymer leaching for sample preparation is very limited, and primarily focused on the detection and analysis of transition metal elements at the ng/g level [2325], as opposed to actinides at much lower concentrations. Glassware is typically avoided for trace level ICP-MS detection because of extraction, or leaching which occurs when in contact with a sample solution [26]. Polymers, despite being more corrosion resistant also leach out a variety of contaminant elements when in contact with corrosive liquids, especially in cases where heat is applied [27]. It has been shown that ultra-trace analyses frequently suffer from contamination of sampling and storage containers [23, 28]. Enhanced detection limits are then of interest for not only the ULB background radiation detection community but also a variety of applications such as ultra-trace ICP-MS sample preparation techniques.

For these reasons, polymer cleaning studies were performed to evaluate contaminant extraction capability for the removal of thorium and uranium from polymeric surfaces. High-density polyethylene (HDPE) was selected as the polymer of interest because of its desirable mechanical attributes, which include toughness, machinability, resistance to abrasion, corrosion, and low coefficient of friction. Because of these properties, HDPE has been identified as a potential ULB material when radioactive element backgrounds can be reduced [29]. The effects of acid composition, concentration, and time duration on the effectiveness of the leaching process were evaluated. Following optimization of the leaching process, mechanical evaluation of leached HDPE polymer material was performed to ensure significant degradation did not occur. Results of this study have demonstrated a simple and effective way to remove surface contamination of U and Th from HDPE. The cleaning process developed and demonstrated should have broad general application to ULB polymer materials and ICP-MS procedures for U and Th analysis at ultra-trace levels.

Experimental details

Materials

Samples of HDPE (McMaster-Carr) were machined into 1.0 cm3 cubes. After machining, the samples were cleaned by sonication in a 5 % solution of Micro90® detergent for 2 h. All labware utilized for preparation or storage of samples, blanks and calibration solutions was cleaned in the same manner and pre-leached three consecutive times for 24 h. Fisher-brand Optima-grade nitric acid diluted to 6 M using >18.2 Mohm water was used for this preliminary leach. Sample cleaning and leaching experiments were performed in a class 10,000 cleanroom environment. Five-milliliter PFA Savillex® vials were purchased specifically for their high purity and chemical compatibility, and were used to contain the leaching test solutions and all samples prepared for analysis. Two HDPE samples were then placed into each pre-leached PFA vial in an effort to normalize any potential differences in surface contamination between HDPE samples. Several different leaching approaches were employed using the reagents shown in Table 1. Testing reagents and concentrations were chosen based on the available literature.
Table 1

Leaching reagents tested

Aqua regia (nitric acid + HCl 1:3 ratio by volume)

6 M Optima nitric acid + 6 M hydrochloric acid

6 M Optima nitric acid

6 M Optima hydrochloric acid

5 % Optima nitric acid by volume

5 % Optima hydrochloric acid by volume

5 % Optima ammonium hydroxide by volume

Deionized water (DIW)

Leaching was performed sequentially with Optima grade acids, leachate aliquots were then removed and analyzed using a self-aspirated Agilent 7700 ICP-MS. Quantitative analysis using the external standard calibration was performed for uranium and thorium to identify the best leach for contaminant removal of these elements. Using the information gathered, the most effective leaching reagents were identified. Following this analysis, process blanks and fully leached specimens were furnace ashed at 1,100 °C in leached quartz crucibles and digested. ICP-MS analysis of these solutions was performed in an effort to determine the level of non-surface U and Th contamination in the polymer.

ICP-MS

Calibration of the Agilent 7700 ICP-MS was performed using natural thorium (m/z 232) and uranium (m/z 238) elemental standards. The instrument was tuned specifically for ultra-trace U, Th detection at the expense of the lower mass range using a six second dwell time, which allows for very high sensitivity of the elements of interest. All solutions were self-aspirated. The calibration performed ranged from ~15–500 ppq 238U and 232Th. While a variety of widely accepted definitions can be found for evaluating detection limits, for this work the detection limit (DL) for each element was calculated as shown in Eq. 1
$$ {\text{DL}} = \frac{{3*B_{\text{STDEV}} }}{{A_{\text{SIGNAL}} - B_{\text{SIGNAL}} }}*[A], $$
(1)
Where BSTDEV is the standard deviation of the blank, ASIGNAL is the analyte intensity in counts per second (CPS) and BSIGNAL is the background signal in CPS. Parameter [A] is the concentration of the analyte of interest. Detection limits were calculated using the 15 ppq standard which was analyzed four times over a 2 months period and were found to be 6 ppq for uranium and 7 ppq for thorium. Blank values were insignificant as averaged values were 16 and 22 % of these concentrations respectively. Calibrations for these elements are shown in Fig. 1
https://static-content.springer.com/image/art%3A10.1007%2Fs10967-012-2301-1/MediaObjects/10967_2012_2301_Fig1_HTML.gif
Fig. 1

Calibration curves for a232Th and b238U calibration showing trace-level detection capability using ICP-MS. Error bars are shown on all points as one standard deviation in the counts per second (CPS) measurement

External contamination introduced during sample preparation would be visible in the Fig. 1 calibration plots as significant deviation from linearity. Even a small amount of contamination introduced would be visible because of the low concentrations of the calibration solutions. Since the calibration curve was shown to be linear without any significant outliers, it can be concluded that the sample preparation techniques utilized did not result in contamination of the sample as it was prepared in the same manner as the calibration standards. The validated cleaning procedure used on the pipette tips and sample vials for the leaching experiment were the same used to prepare the calibration curves shown. Therefore, triplicate 24 h leach using 6 M nitric acid was verified as an appropriate labware leach for performing this study. This preparatory leach was proven to provide a stable contaminant-free baseline for sample vials and pipette tips which would come in contact with the leaching study solutions.

Mechanical testing

Tensile specimens of HDPE were machined according to ASTM D638-99 and tested with the use of an Instron MTS 8800 tension tester at room temperature. The tensile strain rate was set at two inches per minute. Following machining, unleashed samples were tested and compared to samples which had gone through the recommended leaching procedure.

Sample ashing

Leached HDPE cube samples were ashed in acid-leached quartz crucibles using a Thermolyne benchtop muffle furnace. Each crucible was analyzed as a process blank to quantify U and Th backgrounds prior to sample ashing. A two-stage temperature ramp was utilized for the ashing process. Ashing was performed by first heating at 500 °C for 6 h to remove any volatile compounds and carbonize the sample. This was followed by a 4 h ashing step at 1,100 °C. The fully ashed sample was then brought-up into 2 ml of 50 % (v/v) nitric acid. The crucibles were rinsed thoroughly and the final solution was diluted for analysis via ICP-MS.

Results

Surface cleaning method assessment

Various methods of surface cleaning were analyzed, focusing primarily on traditional acid leaching techniques. Figure 2 shows the results of this study. Six molar Optima nitric acid was shown to be the most effective leachate solution for surface contamination removal on HDPE. Greater amounts of thorium were extracted overall compared to uranium. Deionized (DI) water was also analyzed as a control leach to eliminate any concentration-gradient-driven dissolution of thorium or uranium into the leachate solution. No measureable values were obtained with DI water leaching and as such, only detection limits are plotted for these DI water leaches. Because no surface contamination was extracted during the three DI water leaches, a fourth 5 % HCl leach on these samples is representative of a first leach in 5 % HCl as shown. Additionally, by comparing the U and Th results shown in Fig. 2, 5 % HCl as a first leach was shown to be more effective at extracting uranium surface contamination than thorium.
https://static-content.springer.com/image/art%3A10.1007%2Fs10967-012-2301-1/MediaObjects/10967_2012_2301_Fig2_HTML.gif
Fig. 2

Results of leaching trials showing a Thorium concentration in leachate solution with successive leaching and b Uranium concentration in leachate solution with successive leaching. For these figures it should be noted that the leaching reagents are changing with successive leaches. Six molar Optima nitric acid was shown to be the most effective leachate solution for U and Th surface contamination removal. Greater amounts of thorium were extracted overall compared to uranium. Detection limits are plotted for DI water leaches because no surface contamination was extracted during these leaches; as such the fourth 5 % HCl leach on these samples is representative of a first leach in 5 % HCl

Results of leachate effectiveness shown in Fig. 2 can be summarized by the following:
  • 6M nitric acid is the most effective leach for U and Th

  • 5% HCl is more effective at removing U contamination than 5% HNO3

  • DI water has no leaching effect for removal of U or Th on HDPE

  • Ammonium Hydroxide has no leaching effect for removal of U or Th on HDPE

  • 5% HCl and 6M HCl are equally effective at removing Th and U contamination

  • The slight increase shown on leach #4 indicates the HCl matrix was able to remove an additional amount of U and Th contamination not removed by the nitric acid.

An array of different leachate concentrations was tested in an effort to improve cleaning processes and reduce the high costs associated with continuously leaching labware and sample preparation equipment in ultra-pure acids. Results of this work are shown in Fig. 3 with Table 2 describing the leach test identification. Single leaches which varied in duration were performed and analyzed in addition to mixed reagent leaching. For samples which were leached successively, each individual leachate solution was analyzed and the total extracted amount was summed. Results are reported as parts per trillion of thorium and uranium which was extracted into the leachate solution.
https://static-content.springer.com/image/art%3A10.1007%2Fs10967-012-2301-1/MediaObjects/10967_2012_2301_Fig3_HTML.gif
Fig. 3

Cumulative U and Th removed by leachate solutions and reported in parts per trillion. Test ID’s are identified in Table 2. Shaded areas indicate a leachate solution which was an acid mixture. Dashed lines separate acid types and concentration regimes. Specific leaching conditions are identified in Table 2. Six molar nitric acid was shown to be the most effective leaching reagent of those tested

Table 2

Leach test identifications for Fig. 3

ID

Test description

1

24 h in 6 M nitric acid

2

72 h in 6 M nitric acid

3

144 h in 6 M nitric acid

4

400 h in 6 M nitric acid

5

3 × 24 h in 6 M nitric acid

6

72 h in 6 M hydrochloric + 6 M nitric

7

24 h in 6 M hydrochloric acid

8

72 h in 6 M hydrochloric acid

9

144 h in 6 M hydrochloric acid

10

400 h in 6 M hydrochloric acid

11

72 h in aqua regia

12

24 h in 5 % nitric acid

13

3 × 24 h in 5 % nitric acid

14

24 h in 5 % hydrochloric acid

15

3 × 24 h in 5 % nitric/5 % NH4OH/5 % nitric

16

3 × 24 h in DI water

On inspection of the bar chart shown in Fig. 3, several key points can be noted. First, the peaks corresponding to the highest extracted amounts of U and Th were all obtained from leachate solutions containing 6 M nitric acid. Second, higher levels of Th when compared to U were removed from the tested samples. Additionally, the high variability of 238U and 232Th in the leachate solution suggests the presence of these elements on the exterior surfaces of the polymer is also variable. Results of leach duration studies (ID#s 1–4 in Fig. 3) show that leaching in 6 M nitric acid for prolonged periods past 72 h is not significantly impactful which is in agreement with leaching studies performed on polyethylene by Karin et al. [26].

For these reasons, the ideal leaching protocol for effective removal of Th and U surface contamination was identified as a 72 h leach in 6 M nitric acid, followed by a prolonged leach in HCl. And, recalling the results of Fig. 2, there was no measureable difference in extractions performed at 5 % HCl compared to 6 M HCl, making the 5 % HCl leach ideal both for its effectiveness and economical advantage over 6 M HCl. For ultra-trace level analyses, one possible approach would be to store nitric acid leached preparation equipment (pipette tips, vials, containers, etc.) in a solution of 5 % HCl. The 5 % HCl leach could be omitted for less sensitive analyses.

Physical changes in acid leached HDPE

Changes in the appearance of the HDPE samples were observed when leaching was performed in sample solutions 6 (6 M HCl + 6 M HNO3) and 11 (Aqua Regia). Specimens of HDPE submerged in these test solutions exhibited a significant blue discoloration of the polymer. This color change was most likely the result of an interaction with a processing additive incorporated into the HDPE during manufacturing. This effect, coupled with the poor extraction capability of these reagents when compared to 6 M nitric acid, have indicated such acid combinations should not be used for leaching of HDPE polymers due to their inefficiency at extracting Th and U. In an effort to ensure that the recommended leaching procedure did not result in mechanical degradation of the HDPE, evaluation prior and post leach were performed. Tensile specimens were machined from HDPE and tested in the “as machined” state and compared to samples that were leached in accordance with the recommended procedure. Figure 4 shows no measureable degradation in mechanical properties of the polymer was observed. Additionally, no measureable dimensional changes were observed in the polymer samples as a result of leaching.
https://static-content.springer.com/image/art%3A10.1007%2Fs10967-012-2301-1/MediaObjects/10967_2012_2301_Fig4_HTML.gif
Fig. 4

Tensile test results showing the effect of the recommended leaching procedure (72 h in 6 M nitric acid followed by a 24 h leach in 5 % HCl) on the stress versus strain curve of the HDPE tested, Leaching was shown to have a negligible effect on the mechanical properties of the HDPE

Bulk analysis of uranium and thorium in HDPE

An initial bulk contamination study was performed using a leached HDPE sample weighing ~1 g. The sample was ashed and digested to evaluate bulk versus surface contamination of Th and U in HDPE. Bulk polymer results showed uranium and thorium concentrations in the sample to be at 28.5 pg 238U/g HDPE and 6.96 pg 232Th/g HDPE after leaching. When comparing these values to the extracted amounts of U and Th from leaching shown in Fig. 2 (at ~2.4 pg of uranium and 5.6 pg of thorium cumulative from two sample cubes) it appears that majority of contamination is in the bulk material. However, when considering that leaching only removes surface area contamination, the extracted amounts are significant. Additionally, the elevated 238U value obtained from ashing implies there is a more uniform distribution of uranium throughout the sample whereas 232Th appears to be primarily surface contamination.

Conclusion

Results have shown ultraclean acids can provide surface cleaning of HDPE for ULB radiation detection applications. Leaching was shown to be successful at the removal of thorium and uranium and has the potential to be applied to a variety of other contaminants for future testing such as Ra, Pb, and Po in a wide array of polymeric materials.

Six molar nitric acid solutions were shown to extract the most thorium and uranium from the HDPE samples studied. After initial leaching for 72 h in 6 M Optima-grade nitric acid, a 24 h leach using 5 % or 6 M HCl was found to slightly increase contaminant extraction to very similar levels. Tensile testing showed these effective contamination leaching methods did not result in a degradation of mechanical properties or dimensional changes for the HDPE.

As individual leachate solutions both Aqua Regia and a 6 M nitric/6 M HCl mixture were effective at removing U and Th surface contamination, however discoloration of the HDPE was observed after leaching in these solutions, suggesting chemical degradation. For this reason, these solutions are not recommended for use when leaching HDPE materials. From a cost perspective, a 72-hour leach in 6 M nitric acid was shown to be preferred over three, 24-hour leaches. In cases where surface cleanliness is of the utmost importance, this leach should be followed by leaching in 5 % hydrochloric acid.

Future work will include applying acid leaching to a variety of polymeric materials. Additionally, intermediate acid concentrations will also be investigated in an effort to optimize acid cost versus contaminant extraction capability. Because of the potential for a significant variance in surface contamination of samples coming from different processing facilities, a larger study would be needed to accurately assess maximum U, Th surface concentrations which could be removed via the recommended leaching process. Development of reliable, effective protocols for reducing radioactive element background contamination in polymers could ultimately increase measurement sensitivity and therefore be significantly impactful for ULB detection efforts.

Acknowledgments

This work was supported by the PNNL’s Laboratory Directed Research and Development (LDRD) as part of the Ultra-Sensitive Nuclear Measurements Initiative. Pacific Northwest National Laboratory is operated for the US. Department of Energy by Battelle under contract DE-AC06–67RLO 1830.

Copyright information

© Akadémiai Kiadó, Budapest, Hungary 2012