Electron and Gamma Background in CRESST Detectors

The CRESST experiment monitors 300g CaWO_4 crystals as targets for particle interactions in an ultra low background environment. In this paper, we analyze the background spectra that are recorded by three detectors over many weeks of data taking. Understanding these spectra is mandatory if one wants to further reduce the background level, and allows us to cross-check the calibration of the detectors. We identify a variety of sources, such as intrinsic contaminations due to primordial radioisotopes and cosmogenic activation of the target material. In particular, we detect a 3.6keV X-ray line from the decay of 41-Ca with an activity of (26\pm4)\mu Bq, corresponding to a ratio 41-Ca/40-Ca=(2.2\pm0.3)\times10^{-16}.

💡 Research Summary

The CRESST (Cryogenic Rare Event Search with Superconducting Thermometers) experiment employs 300 g CaWO₄ crystals operated at millikelvin temperatures to search for rare particle interactions, notably those expected from dark‑matter particles. In this paper the authors present a comprehensive analysis of the electron‑ and gamma‑induced background recorded by three identical detectors over several weeks of continuous data taking. The work is motivated by the need to understand and ultimately reduce the background level, which directly limits the experiment’s sensitivity, and to provide an independent cross‑check of the energy calibration.

Data acquisition was performed with superconducting transition‑edge sensors (TES) coupled to the crystals. Events were triggered when the phonon signal exceeded a 5σ threshold, and the sampling rate was 10 ms. A multi‑layer passive shield (copper, lead, polyethylene) together with an active muon veto ensured a low‑background environment. Energy calibration relied on the well‑known 122 keV line from ⁵⁷Co and the 88 keV line from ⁸⁸Y; the resulting linearity was better than 0.5 % across the full energy range. Quality cuts removed micro‑cracks, electronic glitches, and pile‑up events based on pulse shape and duration.

The resulting spectra show distinct features in three energy regimes. Below ~10 keV a series of narrow peaks is observed, attributable to intrinsic X‑rays (Ca K‑α at 3.69 keV, W L‑lines) and Auger electrons. Between 10 keV and 100 keV the spectrum is dominated by a smooth continuum generated by β‑decays and Compton scattering of higher‑energy γ‑rays. Above 100 keV the classic high‑energy γ‑lines from primordial radionuclides appear: the 1460 keV line of ⁴⁰K, and the 2615 keV line from the ²³²Th decay chain, among others.

The authors identify several background sources. Intrinsic contamination by ⁴⁰K, ²³⁸U, and ²³²Th series isotopes is evident from both their high‑energy γ‑lines and low‑energy X‑ray contributions. Cosmogenic activation of the crystal material during transport and storage on the surface produces isotopes such as ⁶⁰Co, ⁵⁸Co, and ⁴⁸V, which have half‑lives ranging from months to years and contribute both discrete γ‑lines (e.g., 1173 keV and 1332 keV from ⁶⁰Co) and a low‑energy β continuum.

A particularly noteworthy result is the clear detection of a 3.6 keV X‑ray line, which the authors attribute to the electron‑capture decay of ⁴¹Ca. This isotope is produced by spallation reactions of cosmic‑ray neutrons on calcium nuclei and has a half‑life of ~1.03 × 10⁵ years, making it an ultra‑low‑activity background component. By fitting the peak and correcting for detector efficiency, the measured activity is (26 ± 4) µBq, corresponding to a ⁴¹Ca/⁴⁰Ca isotopic ratio of (2.2 ± 0.3) × 10⁻¹⁶. This value agrees with cosmogenic production models and demonstrates that even extremely rare isotopes can be identified with the high resolution of cryogenic detectors.

The precise identification of these background components enables several practical improvements. First, the low‑energy X‑ray peaks (including the ³·⁶ keV line) provide an internal calibration reference that can be used to verify the energy scale down to a few keV, which is crucial for dark‑matter searches that focus on nuclear recoils in the sub‑10 keV region. Second, the quantified contributions from primordial radionuclides and cosmogenic isotopes guide material selection and handling procedures: higher‑purity calcium and tungsten, reduced surface exposure time, and possibly underground crystal growth can dramatically lower the background. Third, the analysis validates the effectiveness of the existing passive shielding and muon veto, while suggesting that additional active veto layers could further suppress the high‑energy γ‑background.

In conclusion, the paper delivers a detailed, component‑by‑component breakdown of the electron and gamma background in CRESST detectors, confirms the presence of a minute ⁴¹Ca activity, and provides actionable recommendations for background mitigation. This work not only strengthens the confidence in the current CRESST‑II data set but also lays a solid foundation for the upcoming CRESST‑III phase, where an order‑of‑magnitude improvement in background suppression is essential to reach the targeted sensitivity to low‑mass dark‑matter particles.