Large-Scale Production of Radiopure 135Xe from Bremsstrahlung γ-Irradiation of Solid Xenon Difluoride

This article describes the use of bremsstrahlung radiation to induce¹³⁶ Xe(γ,p)¹³⁵I reactions, followed by milking of ¹³⁵Xe from β-decay of the ¹³⁵I product. To increase Xe density in irradiated samples, we used solid xenon difluoride, resulting in a yield of 80 nCi ¹³⁵I and 26 mCi ¹³⁵Xe from 122 mg XeF₂ after irradiation of 21.9 MeV over 30 min. Using XeF₂, we were able to separate unreacted xenon from the ¹³⁵I product via aqueous sodium hydroxide decomposition of XeF₂, followed by removal of evolved gaseous xenon by cryotrapping with liquid nitrogen. After decay, radiopure ¹³⁵Xe was recovered by cryotrapping.

Xenon-135, or ¹³⁵Xe, is required to calibrate the detectors that passively monitor the atmosphere for global enforcement of the 1963 Nuclear-Test-Ban Treaty. Radioactive xenon in the atmosphere can indicate nuclear explosions or testing; it is a clear marker of underground testing (because the unreactive noble gas escapes and persists after a test). Indeed, monitoring ¹³³Xe and ¹³⁵Xe ratios offered the most reliable proof of North Korea’s 2006 underground nuclear test, providing a fission source term in contrast with seismic data that could have been produced with nonnuclear explosives.^1,2 Optimized purification of xenon from atmospheric sampling has been demonstrated using the automated radioxenon sampler-analyzer (ARSA) since 1999. ARSA was developed by the US Department of Energy’s Office of Nonproliferation and the Pacific Northwest National Laboratory. Engineers have successfully purified and concentrated xenon from the atmosphere for analysis of radioxenon used to detect nuclear activity.

In experimental nuclear physics, a two-body nuclear reaction can be expressed as A(a,b)B, where, by convention, A represents the target nucleus; a is the incident particle; b is a lighter ejected particle, known as the ejectile; and B is a heavier ejected particle, known as the recoil. ¹³⁴Xe(n,γ)¹³⁵Xe transitions after exposure of low-density gaseous xenon, in which a neutron is incident on a ¹³⁴Xe nucleus, produce valuable ¹³⁵Xe samples for calibrating the worldwide detectors. These samples, however, contain mostly ¹³⁴Xe, which complicates the calibration process. Not only would radiopure and carrier-free ¹³⁵Xe samples improve the existing state of the art, but using an alternative, denser form of Xe would allow for production of greater yields.

¹³⁶Xe(γ,p)¹³⁵I reactions have been described as a way to produce ¹³⁵Xe. Horkley et al. recently reported production of 592 Bq of ¹³⁵Xe after 6.5-h irradiation at 21.8 MeV.³ These experiments produced ¹³⁵Xe without detectable ¹³³Xe; however, the yields were 10³ times lower than would be required for practical use. If using a similar gamma source, concentrating the xenon substrate solves this limitation.

Xenon difluoride, or XeF₂, is a colorless salt produced from the reaction of xenon and difluorine. These gases react when energy is added. XeF₂ has been synthesized from irradiation with ultraviolet (UV) light, heat, γ-rays, and other energy sources.⁴ Today XeF₂ is typically produced via UV irradiation of a xenon fluorine mixture in a 2:1 ratio with catalytic 1 mol % hydrogen fluoride,⁵ which forms pure crystals. Because of its volatility and reactivity, XeF₂ is best stored cold and in a sealed polyethylene or fluoropolymer container. Although xenon is a noble gas and therefore tends to be chemically inert, xenon difluoride is reactive; it will react with water but not rapidly, and it requires the presence of base to reliably dissociate and evolve xenon gas.⁶ XeF₂ is used as a niche fluorinating agent for research purposes⁷ but has limited industrial use synthetically. XeF₂ is available commercially from common chemical suppliers and can be purchased for less than $200 per gram, although not in isotopically pure forms. Furthermore, purifying xenon from XeF₂ is simple by using liquid nitrogen for cryo-condensation.

Efforts to optimize the purification of xenon from atmospheric sampling have demonstrated reliable sequestration of radioxenon produced from ¹³⁶Xe(γ,p)¹³⁵I reactions and subsequent ¹³⁵I β-decay. In this article, we report a proof of concept for generating ¹³⁵Xe in large yields from a solid xenon-containing substrate, followed by purification and trapping of radiopure ¹³⁵Xe.

Linear Accelerator Facility

We conducted experiments in APL’s linear accelerator facility, a ~4,000-ft² laboratory that consists of a ~1,000-ft² heavily shielded accelerator hall, a ~600-ft² heavily shielded experimental hall, a ~620-ft² radiochemistry laboratory, and a ~520-ft² experiment staging area. The accelerator and experimental halls are both partially underground with walls and ceilings consisting of ~3 ft of concrete and 1.7 ft of steel, providing ample radiation shielding for radiation-producing machines and large radioactive sources. The accelerator is capable of delivering electron pulses with operator-selectable widths between 0.5 and 4 ms at a repetition rate up to 115 Hz. The maximum charge per pulse ranges from 40 nC for 0.5-ms pulses to 350 nC for 4-ms pulses. The electron energy can be varied from ~5 to 25 MeV with an energy resolution of 5%, which can be reduced using collimators or slits. The S-band standing wave radio frequency linear accelerator, which operates at ~2.8 GHz, was used to accelerate 21.9-MeV electrons that generate bremsstrahlung γ-radiation using a 2-mm tungsten source. With the electron linear accelerator’s energy range and versatility, we were able to develop a novel solution for synthesizing radiopure ¹³⁵Xe.

Experimental Methods

We placed xenon difluoride crystals (200 mg) into a 25-mL single-neck borosilicate glass Schlenk reaction tube fitted with a rubber septum. Next we submerged the sealed reaction tube in an ice water bath and then placed it directly in the γ-beam. This setup is pictured in Figure 1.

We fit the sealed reaction tube containing irradiated XeF₂ to the apparatus pictured in Figure 2. The apparatus comprised the sealed reaction tube connected to a drying trap that facilitated visualization of moisture trapping. The drying trap connected to a stainless steel U-tube, also shown in Figure 2, and was fitted with a manual valve on both ends and a gauge to monitor internal pressure. We introduced into the reaction tube a steady stream of ultrahigh-purity nitrogen from a cylinder. Next, we opened the tube, drying tube, and reaction tube valves and then submerged the U-tube in liquid nitrogen before adding aqueous sodium hydroxide (1 N, 3 mL) to the XeF₂ crystals. Nitrogen flowed into the reaction solution to sparge the mixture and remove the entirety of dissolved xenon, as shown in Figure 3. We continued to purge the system with nitrogen for 1 min after we could no longer see xenon evolution, and then we shut off the gaseous nitrogen flow. Finally, we closed the reaction tube, drying tube, and tube valves and removed the U-tube from the liquid nitrogen bath, allowing the tube to warm to room temperature. Figure 4 shows the XeF₂-containing reaction tube at various stages during these experiments.

We piloted this procedure with unirradiated XeF₂ and examined the collected xenon by residual gas analysis, which is a form of mass spectrometry. During mass spectrometry, the mass-to-charge ratio (m/z) of an ion is measured, typically by manipulating the ion’s trajectory using a magnetic field.⁸ The sensitivity of mass spectrometry techniques is such that individual ions of a given mass-to-charge ratio can be counted. This analysis revealed xenon isotopes ranging from 126 to 136 Da, similar to the natural isotopic distribution for xenon, as shown in Figure 5. An increase in ion counts can be observed in this range for the xenon sample, shown in blue.

Irradiation of Xenon Difluoride

We sought to irradiate xenon difluoride crystals and determine the yield of ¹³⁵I and ¹³⁵Xe produced from a solid substrate. In a first-pass experiment, we added 100 mg XeF₂ to a borosilicate Schlenk tube. Processes to synthesize XeF₂ involve adding energy during the reaction, including γ-irradiation, and we did not know whether adding high amounts of ionizing radiation could reverse the reaction and produce hazardous gases.⁹ Therefore, we fit the Schlenk tube to a gas scrubber to remove any produced hydrogen fluoride or fluorine vapors with aqueous 3N sodium hydroxide. This precaution turned out to be unnecessary, however, as we observed no bubbling during irradiation, and the XeF₂ crystals appeared intact after irradiation for 15 min at 21.9 MeV. Longer irradiations, at 1 h, resulted in visible sublimation of the XeF₂ substrate, so we performed future experiments at 0°C by using an ice water bath. With the temperature held constant at 0°C, we observed no sublimation, and the XeF₂ crystals remained intact and in place at the bottom of the reaction tube.

We measured the production yields of ¹³⁵I and ¹³⁵Xe at multiple energies by bombarding a 200-mg sample of XeF₂ with bremsstrahlung radiation. Measurements were taken at 16-, 19-, and 22-MeV end-point energies at 15-min irradiation intervals. Electrons were accelerated using the linear accelerator facility’s 25-MeV linear accelerator (Figure 6), and bremsstrahlung radiation was generated by impinging a 2-mm-thick tungsten radiator. Also known as braking radiation, bremsstrahlung results from electrons decelerating in a medium.^11,12 A 2.54-cm-thick block of aluminum served as a beamstop. A simulated bremsstrahlung spectrum for an incident 22-MeV electron beam is shown in Figure 7.

Purification of Xenon from Xenon Difluoride

We hypothesized that XeF₂ could be rapidly dissociated into xenon gas and fluoride salts upon addition of aqueous sodium hydroxide. Indeed, adding 3 mL 1N NaOH solution to the XeF₂ crystals caused vigorous bubbling in our initial tests with unirradiated material. We assembled an apparatus to remove xenon after irradiation by using positive flow of nitrogen gas from a cylinder into the sample-containing Schlenk tube, followed by drying of the flowed gases through anhydrous calcium sulfate desiccant and condensation of xenon gas into a stainless steel trap made from ¼-in. stainless steel tubing using liquid nitrogen. Xenon was cryotrapped from unirradiated XeF₂ using our apparatus, and successful xenon recovery was validated by residual gas analysis mass spectrometry.

The experimental workflow was repeated with irradiated XeF₂ and iodine was successfully maintained in the aqueous sodium hydroxide solution. The isotopes of interest, ¹³⁵I and ¹³⁵Xe, have half-lives of 6.9 ± 0.4 h¹³ and 9.13 ± 0.5 h,¹⁴ respectively. Furthermore, it can be extrapolated that ¹³⁵I β-decays exclusively to ¹³⁵Xe based on the energies reported for this transition.¹⁵ Characteristic γ-decay energies associated with these radio­isotopes were measured by analyzing the cryotrapped gases using a high-purity germanium detector (HPGe), which is a semiconductor γ-ray detector.¹¹ We observed prominent γ-ray decays of energy Eγ = 1,260.409, 1,131.511, and 1,457.56 keV, as shown in Figure 8. Decay curves, shown in Figure 9, were constrained for each isotope and are within error bars of known values. The production yields for ¹³⁵I and ¹³⁵Xe are shown in Figure 10. We determined that ¹³⁵Xe was successfully purified from the ¹³⁵I sample. After waiting for ~16 h, we repeated the procedure to sparge the solution and induce cryotrapping of the ¹³⁵Xe resulting from β-decay of the ¹³⁵I product during that period. Radiopure ¹³⁵Xe was recovered.

An additional insight resulted from these measurements. Currently, the experimental cross sections of the ¹³⁶Xe(γ,n)¹³⁵Xe and ¹³⁶Xe(γ,p)¹³⁵I reactions, which describe the probability of these reactions taking place, have not been measured. As such, theoretical cross sections are used to predict the production yields of the ¹³⁵I and ¹³⁵Xe isotopes. Our measured ¹³⁵I production yield suggests a large discrepancy between the theoretical and experimental cross sections. Further measurements are needed to clarify this discrepancy and to constrain the experimental cross sections for these reactions; this work would be beneficial to the nuclear data community.

Our hypothesis for how ¹³⁵Xe could be purified from the irradiated solid XeF₂ hinges on the passive evolution of gaseous xenon from aqueous solution. The overall radiochemical transition from ¹³⁶Xe to ¹³⁵I is shown in Figure 11. As the ¹³⁵I is likely stable as a salt in aqueous solution, the remaining xenon can be removed at this point by sparging the aqueous solution with an inert gas, such as nitrogen. The remaining aqueous ¹³⁵I will decay into ¹³⁵Xe with a half-life of 6.6 h and then can be removed by sparging the solution with gaseous nitrogen. Xenon’s boiling point is 165 K, compared with nitrogen’s boiling point of 77.3 K; therefore, the evolved ¹³⁵Xe can be trapped using liquid nitrogen.

The proposed chemistry in Figure 11 allows ¹³⁵Xe to be selectively purified if radiopure ¹³⁶XeF₂ is used because the unconverted xenon is removed after the addition of aqueous sodium hydroxide (NaOH) followed by sparging of the solution with nitrogen. The removal of evolved ¹³⁵Xe after a waiting period through cryotrapping leaves behind the ¹³⁵I as an aqueous salt. The purity of the milked ¹³⁵Xe is measured using HPGe γ-spectroscopy, demonstrating the effectiveness of this approach for large-scale ¹³⁵Xe production. An example of the HPGe γ counting setup is shown in Figure 12.

These methods could be applied to a more refined production process using a custom-built and engineered apparatus. The techniques involved are simple and rely on separation of undesired xenon isotopes by removing them before the majority of the ¹³⁵I decay. If inert conditions are maintained through the use of a nitrogen atmosphere, using liquid nitrogen to cryotrap the radioactive xenon product gas is a reliable approach to separating ¹³⁵Xe from the mixture without risking chemical release as a result of system overpressure.

Using XeF₂ as a substrate for radiochemical production of ¹³⁵Xe is a viable strategy to generate calibration samples for the comprehensive Nuclear Test Ban Treaty’s atmospheric xenon detectors. Not only are production yields sufficient, but the ability to chemically separate the produced ¹³⁵I allows for isolation of radiopure ¹³⁵Xe and removes the presence of problematic xenon isotopes. Cryo-condensation of these undesirable xenon isotopes allows for their safe handling and recycling, as well as bottling of radiopure ¹³⁵Xe for future sample preparation.

Future directions to refine this approach would involve designing a ¹³⁵Xe separation apparatus and using radiopure ¹³⁶XeF₂. An apparatus for ¹³⁵Xe evolution and separation would require the use of an inert atmospheric environment, as previously stated, so the ability to load the irradiated XeF₂ material and purge the system with gaseous nitrogen is required, followed by the ability to add aqueous sodium hydroxide solution to the XeF₂ and sparge the resulting solution without opening the system to the atmosphere. Ideally, the evolved Xe that contains undesired ¹³⁶Xe from the sample would be cryotrapped and recycled into the synthesis of ¹³⁶XeF₂. The apparatus should allow for an additional or replacement cryotrap vessel so that pure ¹³⁶Xe can be collected after the ¹³⁵I decay process without opening the system to the atmosphere.

Radiopure ¹³⁶XeF₂ can be synthesized through exposing UV light to a mixture of radiopure gaseous ¹³⁶Xe, available from chemical vendors in high purity, and fluorine gas, which is also commercially available. Although gaseous fluorine is highly corrosive and toxic, other researchers have described the synthesis of XeF₂ in a controlled laboratory setting.⁴ APL researchers could reproduce these settings in their laboratory or contract chemical manufacturers to do so (though the latter option would be more expensive).

We were able to demonstrate a novel production pathway for the synthesis of radiopure ¹³⁵Xe through three innovations: the milking of ¹³⁵Xe via the ¹³⁶Xe(γ,p)¹³⁵I reaction (where a XeF₂ target was bombarded with bremsstrahlung radiation generated by the 25-MV linear accelerator), the use of a solid XeF₂ target instead of a liquid solution (which increases the density of Xe in the target), and the chemical extraction of ¹³⁵I from the irradiated XeF₂ target. If implemented, our production pathway for radiopure ¹³⁵Xe could produce sufficient radioxenon to supply calibration material to the entirety of the worldwide network of detectors that passively sample radioxenon from the atmosphere.