In-situ measurement of the scintillation light attenuation in liquid argon in the Gerda experiment

https://doi.org/10.1016/j.nima.2019.163059Get rights and content

Abstract

The Gerda experiment searches for the neutrinoless double beta (0νββ) decay in 76Ge in order to probe whether the neutrino is a Majorana particle and to shed light on the neutrino mass ordering. For investigating such a rare decay it is necessary to minimize the background of the experiment. In Phase II of the Gerda experiment the scintillation light of liquid argon (LAr) is used as an additional background veto. In order to estimate the efficiency of such a LAr veto it has to be known how far the scintillation light, which peaks at 128 nm, can travel within the LAr. A dedicated setup was built to measure the attenuation length of the scintillation light in the LAr in-situ within the cryostat of Gerda. The setup is composed of a stainless steel housing with a photomultiplier tube (PMT) at one side and a moveable 90Sr source at the other side to measure the light intensity at different distances between source and PMT. Furthermore, a sophisticated simulation was developed in order to determine the solid angle correction as well as the background for this measurement. The analysis results in an absorption length of 15.8±0.7(stat)3.2+1.5(syst)cm under the assumption of a scattering length of 70 cm at 128 nm. The obtained value of the absorption length is specific for the LAr in GERDA at the time of the measurement.

Introduction

The germanium detector array (Gerda) is an experiment located in the underground laboratory Laboratori Nazionali del Gran Sasso (LNGS) and searches for the neutrinoless double beta (0νββ) decay in 76Ge [1]. Gerda operates isotopically enriched high purity germanium (HPGe) detectors in liquid argon (LAr), which serves as a cooling and shielding against external radiation.

The current limit of Gerda on the half live of the 0νββ decay of 76Ge is T12>0.91026yr (90% C.L.) [2]. The search for such a rare decay requires to minimize the background of the experiment. One of the major improvements in Gerda Phase II is an additional veto system surrounding the HPGe detector array using the LAr scintillation light. Excess energy that gets deposited in the LAr will trigger the scintillation light providing a signal to further suppress background. This includes intrinsic impurities in the LAr that could mimic the double beta decay as well as nuclides produced by cosmogenic activation inside the detectors and other natural abundant nuclides inside the surrounding material like detector holders and cables. For this purpose, the LAr cryostat in Gerda was instrumented with a system consisting of a combination of reflectors, wavelength shifters (WLS) and photomultiplier tubes (PMT).

The scintillation light in LAr is created by the emission of a de-excitation photon of 128 nm from either a singlet or a triplet state of an excited LAr molecule. The ratio between the population of singlet and triplet state is 0.3 for electrons. While the transition of the singlet state is allowed (4–7 ns), the triplet state is forbidden (1.0–1.7μs), resulting in different decay times [3].

In case the excited LAr molecule collides with impurities, such as oxygen or nitrogen, it can de-excite non-radiatively, decreasing the light yield as well as the lifetime of the triplet state. Furthermore, oxygen can absorb scintillation photons directly, which also decreases the light yield as well as the absorption length even more.

Consequently, the scintillation light cannot propagate infinitely inside LAr but is attenuated due to present impurities [4], [5], [6], which causes a limit on the effective active volume of the LAr veto. In order to estimate its efficiency the attenuation length of the scintillation light was measured inside the Gerda cryostat.

Since the attenuation length is just one of the key parameters and other important values are still poorly known, the estimation of the LAr efficiency is investigated by a subgroup of Gerda and is not covered in this work.

An in-situ measurement has been performed because the attenuation is strongly dependent on the specific impurities present in the Gerda LAr cryostat. The results of light attenuation measurements of other experiments cannot be applied to Gerda due to their different impurity content.

In the experiment it is not possible to disentangle scattering αscat and absorption αabs lengths and only the attenuation length 1αatt=1αabs+1αscat is measured. For the implementation of αabs and αscat into the simulation, αabs is determined under different hypotheses of αscat. This is done because αscat is less dependent on impurities in LAr, there are better estimates of it and assuming typical values, it is not the dominant effect in Gerda.

Section snippets

Experimental setup

For the measurement of the attenuation of the scintillation light in the LAr inside Gerda a dedicated setup was designed that could be submerged directly into the LAr cryostat. A CAD drawing of the setup is shown in Fig. 1.

The setup consists of a steel housing with a PMT mounted on one end and a 7 kBq 90Sr source in front of the PMT, which is held by movable steel rods. The total length of the setup is 1 m and the source can be moved by the steel rods in order to measure the scintillation light

Data taking and signal reconstruction

Pulse traces of the PMT were recorded with a 14-bit Fast Analog to Digital Converter (FADC) board ( Struck, SIS3301) with a sampling frequency of 100 MHz and a trace length of 131 072 samples (approx. 1.3 ms).

All traces were saved for each source position to allow for a complex offline analysis. In a first step, all these traces were cleaned from a periodic interference of 20 kHz caused by the operation of the stepper motor.

The motor interference would falsify the baseline determination before

Simulation studies

A Monte Carlo simulation of the setup was performed in order to model the experimental results and to test the analysis procedure on simulated data.

The absorption length is implemented in a wavelength dependent way following reference [12] and scaled to achieve a certain absorption length at 128 nm in the simulation. However, the measurement is not sensitive to the wavelength since the scintillation spectrum is shifted by the WLS and accumulated by the PMT. Because of this it is not possible to

Data analysis

In this section the analysis technique successfully evaluated for the simulation in Section 4.5 is applied to the measured data. For each reflectivity assumption at 128 nm a respective solid angle correction is obtained. Following, the data are analyzed with each solid angle correction and the combined fit f(Δd) is applied according to Eq. (4) using the Cherenkov background fit of the simulation.

Conclusion and outlook

An attenuation length measurement was performed in the Gerda LAr cryostat with a dedicated setup. The attenuation length has been extracted from the experimental data using detailed simulations of the setup to take into account Cherenkov background, geometrical acceptance of the PMT as well as details of the light propagation. The simulation framework was further used to cross-check the subsequent analysis before applying it to the measured data. The input parameters which lead to the best

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

This work was supported in part by grants from BMBF, Germany, DFG, Germany, INFN, Italy, MPG, Germany, NCN, Poland, RFBR, Russia and SNF, Switzerland . The authors thank the workshop of TU Dresden for building the setup and the colleagues and the workshop of the MPIK in Heidelberg for the opportunity to test the setup in liquid nitrogen. We thank in particular Bernhard Schwingenheuer for his continuous support and help during the preparation and execution of the measurement as well as the

References (25)

  • BellamyE.H.

    Nucl. Instrum. Methods Phys. Res. A

    (1994)
  • SeidelG.M.

    Nucl. Instrum. Methods Phys. Res. A

    (2002)
  • IshidaN.

    Nucl. Instrum. Methods Phys. Res. A

    (1997)
  • Bideau-MehuA.

    J. Quant. Spectrosc. Radiat.

    (1981)
  • BenettiP.

    Nucl. Instrum. Methods Phys. Res. A

    (2007)
  • AgostiniM.

    Phys. Rev. Lett.

    (2018)
  • A.J. Zsigmond, http://dx.doi.org/10.5281/zenodo.1287603...
  • AmslerC.

    J. Instrum.

    (2008)
  • JonesB.J.P.

    J. Instrum.

    (2013)
  • AcciarriR.

    J. Instrum.

    (2010)
  • JonesB.J.P.

    J. Instrum.

    (2013)
  • The Lund/LBNL Nuclear Data Search...
  • 1

    Present Address: Laboratório de Instrumentação e Física Experimental de Partículas Lisboa Portugal.

    2

    Present Address: Nuclear Science Division, Lawrence Berkeley National Laboratory, Berkeley California 94720 USA.

    View full text