Cosmogenic activation of sodium iodide

The production of radioactive isotopes by interactions of cosmic-ray particles with sodium iodide (NaI) crystals can produce radioactive backgrounds in detectors used to search for rare events. Through controlled irradiation of NaI crystals with a neutron beam that matches the cosmic-ray neutron spectrum, followed by direct counting and fitting the resulting spectrum across a broad range of energies, we determined the integrated production rate of several long-lived radioisotopes. The measurements were then extrapolated to determine the sea-level cosmogenic neutron activation rate, including the first experimental determination of the tritium production rate: $(80\pm 21)\mathrm{atoms}/\mathrm{kg}/\mathrm{day}$. These results will help constrain background estimates and determine the maximum time that NaI-based detectors can remain unshielded above ground before cosmogenic backgrounds impact the sensitivity of next-generation experiments.

Figure 1

Experimental measurement [21] and geant4 model estimates (see text for details) of neutron-induced tritium production in sodium (solid curves) and iodine (dotted curves).

Figure 2

Experimental measurements [36, 37, 38] and geant4 model estimates of neutron-induced ${}^{22}\mathrm{Na}$ production in sodium (continuous curves) and iodine (dotted curves). Measurements of the proton-induced cross section (gray markers) [38] are also shown for reference.

Figure 3

Experimental measurements [39, 40] and geant4 model estimates (continuous curves) of neutron-induced ${}^{125}\mathrm{I}$ production in iodine. Measurements of the proton-induced cross section (gray markers) [38] are also shown for reference.

Figure 4

Picture of the three NaI crystals mounted on sliding acrylic holders being aligned with the neutron beam at the LANSCE ICE-HOUSE II facility. The beam direction is angled out of the page and passes through the cylindrical fission chamber, seen on the right, before passing through a crystal.

Figure 5

Comparison of the LANSCE 4FP30R/ICE II neutron beam with sea-level cosmic-ray neutrons. The data points and left vertical axis show the number of neutrons measured by the fission chamber during the beam exposure for each crystal. Vertical error bars shown are statistical only, while the shaded region indicates the systematic uncertainty from the fission cross section (see the main text for discussion of systematic uncertainties). For comparison, the red continuous line and the right vertical axis show the reference cosmic-ray neutron flux at sea level for New York City during the midpoint of solar modulation [52].

Figure 6

Experimental measurements (circles) [45, 47, 49] and evaluations (squares) [46, 48, 50, 51] of the ${}^{238}\mathrm{U}(\mathrm{n},\mathrm{f})$ cross section. The cross section assumed by the LANSCE facility to convert the fission chamber counts to a total neutron fluence is shown by the black line, with the shaded gray band indicating the assumed uncertainty (see Sec. 3c for details).

Figure 7

A geant4 rendering showing the NaI crystal and the fission chamber on the LANSCE beam. The horizontal distance between the crystal and the fission chamber is not shown to scale due to space limitations.

Figure 8

Diagram of the experimental setup at Wright Laboratory. The irradiated NaI crystal’s fused silica window was optically coupled to a PMT, which was contained in a light-tight aluminum capsule. Measurements were performed within an outer lead castle (4 in. thick bricks) and an inner aluminum box (0.25 in. thick) to shield the detector from environmental backgrounds.

Figure 9

The simulated spectra of uncorrected energy deposits (red) and after applying the light yield nonlinearity model (black) in the crystal for ${}^{125m}\mathrm{Te}$ (solid lines) and ${}^{113m}\mathrm{In}$ (dashed lines). Note that the inclusion of the light yield nonlinearity shifts the peaks in opposite directions as described further in the text.

Figure 10

The experimental measurement of the induced activity at February 2021 (blue) and the result of the month-by-month fitting (black) at PMT voltages of (a) $-1500$, (b) $-1000$, and (c) $-800\mathrm{V}$. The most important isotopes contributing to the fit in each energy range are also shown (dashed lines).

Figure 11

The time dependent activity returned by the month-by-month fitting. The dashed lines indicate the best fits to an exponential decay. The deduced half-lives are compared with literature values in Fig. 12.

Figure 12

The half-lives for the isotopes produced in the NaI, as determined by an exponential fit to their monthly activities returned by the fit (black), along with the fit uncertainties. The nuclear data values are also shown (red).

Figure 13

The data across the entire energy range for the December 2020 (blue), March 2021 (green), Jun 2021 (red), September 2021 (orange), and December 2021 (purple) measurements, and the simultaneous fit at those times with a corresponding color and a lighter shade. The spectral features are labeled with the identity of the isotope primarily responsible for each feature.

Figure 14

The time dependence of the ${}^{125m}\mathrm{Te}$ activity. The data (red) are well described using a model (blue) with decays from initial ${}^{125m}\mathrm{Te}$ production as well as feeding from ${}^{125}\mathrm{Sb}$. The individual exponential components due to ${}^{125m}\mathrm{Te}$ (solid black) and ${}^{125}\mathrm{Sb}$ (dashed black) are shown for comparison.

Figure 15

Comparison of the results from this work (black circles) to previously measured comsogenic production rates (colored circles) in ANAIS-112 [15, 19], DM-Ice17 [20], and COSINE-100 [16]. Also shown (colored triangles) are averaged calculations of tritium production from ANAIS [15] and estimates from the ACTIVIA [15, 63, 64] semiempirical model. See Sec. 8 for details.