Direct Detection
Dark Matter Direct Detection
Only 5% of our Universe consists of ordinary matter - like electrons, protons, and neutrons. What about the rest? Direct dark matter detection tries to answer this question trying to find the evidence of the interaction between the elusive dark matter particles and atomic nuclei or electrons from detector targets. These experiments look for events which are rare, and difficult to distinguish from those induced by natural radioactivity, which contaminates detector materials and the surroundings. They are often operated at underground facilities, which are naturally shielded from cosmic rays, and consist of kg to multi-tonne-size targets instrumented in order to be able to detect the recoil of ordinary particles after they are hit by dark matter.
The Double-Phase Time-Projection Chamber Technology
The recoil between dark matter and atomic nuclei or electrons must be carefully reconstructed, as the energy carried by the recoiling particle can be used to tell whether dark matter hit the detector, and know something about its mass. In general, this recoiling particle (let’s say a nucleus) travels in the target medium and may excite its atoms, ionize them or transfer them energy which goes into heating. The more of these signals are reconstructed by an experiment, the better its capability to identify and measure the energy of the original scattering. By counting how many events one observes as a function of this energy, one can know something about the mass and interaction probability of dark matter with ordinary matter.
DarkSide uses the dual-phase time projection chamber technology (TPC) to do so. The detector can be seen as a large barrel filled up with liquid argon, instrumented at its top and bottom endcaps with photodetectors, while its lateral faces are internally coated with a reflective layer. The barrel isn’t fully filled with liquid, but there is a small (few mm) layer of gaseous argon at its top, separated from the liquid by a finely-grained metallic grid. This detector technology is able to measure scintillation and ionization signals, due to the de-excitation and ionization of argon atoms, which live within the liquid phase. Once a dark matter particle scatters off a nucleus, for example, the nucleus will travel in the liquid, and lose its energy by exciting or ionising other argon atoms. Scintillation light is produced in the form of photons emitted immediately after the scattering, and is detected by the photodetectors (possibly after having been reflected by the lateral faces). Argon ion-electron pairs are produced due to ionization: electrons would naturally recombine with other argon ions in the liquid, but a first electric field (“drift field”) is applied to the TPC and makes them travel upwards. Once they reach the metallic grid, they experience a second electric field which is able to extract them from the liquid phase, and accelerate them towards the top endcap of the TPC (anode). This acceleration is abrupt, and induces an avalanche multiplication of electrons: by measuring the light emitted in this process (“electroluminescence light”) one is able to reconstruct number of ionization electrons produced by the dark matter scattering.
These two signals are called S1 (scintillation) and S2 (ionization), and they are delayed proportionally to the vertical position of the scattering: the closer to the top endcap, the closer S1 and S2 are in time. This allows to reconstruct the vertical coordinate (“z”) of the scattering; the two horizontal coordinates (“x” and “y”) are instead measured by comparing the signal collected by the matrix of photodetectors placed on the TPC endcaps. Reconstructing x, y, and z is crucial to be able to discriminate events due to dark matter from events due to natural radioactivity, i.e. background events which are induced by particles emitted in radioactive decays happening in the detector materials or in its surroundings.
The usage of liquid argon enables also the capability to discriminate nuclear and electron recoil events, by analysing how the S1 signal evolves in time. This technique is called pulse-shape discrimination (PSD), and exploits the mechanisms which rule the production of scintillation light: signals due to nuclear recoils tend to be emitted in argon with characteristic times of a few nanoseconds, while signals due to electron recoils tend to be 1000 times slower (few microseconds). By measuring how “prompt” the S1 signal is, one is able to discriminate fast and slow signals, and in turn nuclear and electron recoils - which makes it easier to distinguish the interaction of dark matter with a nucleus from the interaction of a radioactivity-induced electron with another atomic electron.
DarkSide-50
The DarkSide-50 TPC at the INFN Laboratori Nazionali del Gran Sasso (LNGS) consists of a ~50 kg volume of liquid argon, extracted from underground sources and as such depleted from radioactive isotopes produced in argon exposed to cosmic rays. The DarkSide-50 TPC is a cylinder of 36.5 cm height and diameter, instrumented with 38 photomultiplier tubes (19 per endcap). The TPC is surrounded by an additional liquid scintillator detector (Liquid Scintillator Veto, LSV), filled with 30 t of liquid boron, which is chosen to maximise the probability of detecting possible interaction between neutrons and argon nuclei, a signal which would mimick the dark matter-nucleus interaction. The LSV is itself surrounded by a water-based Cherenkov detector (WCV) which is used to reject cosmic ray muons. The LSV and WCV are called veto detectors, as they are used to veto events: whenever a signal is reconstructed in the TPC and in the LSV and/or WCV, the event is rejected as it’s most probably not due to a dark matter interaction.
DarkSide-50 took data from 2013 to 2018, and analysed them by looking for high- and low-mass dark matter. The former search is a so-called (almost) “background-free” approach, in which the combination of the PSD parameter f90 (fraction of S1 signal collected in the first 90 ns from a scattering event) and S1 vs S2 reconstruction is used to suppress background events to less than 0.1 events in the full dataset; its results were published in Phys. Rev. D 98, 102006 (2018). The latter instead relies only on the S2 signal (“S2-only”), as the nuclear recoil energy for dark matter particles of mass close to 1 GeV is insufficient for the S1 signal to be reconstructed efficiently; its results were published originally in Phys. Rev. Lett. 121, 081307 (2018), and were recently reanalysed using a larger dataset and improved detector calibration. The low-mass dark matter search results from DarkSide-50 proved the effectiveness of the S2-only technique in noble liquid TPCs, and provide the best results to date for dark matter between ~1 GeV and ~4 GeV of mass.
DarkSide-20k
The Global Argon Dark Matter collaboration is now building a larger version of DarkSide, using a 20 t liquid argon target. This new detector, DarkSide-20k, will operate at LNGS and use thousands silicon photomultipliers as light detectors. This improved detector technology brings new challenges for the reconstruction of signals in the TPC, which will make use of advanced signal processing and machine learning techniques to extract and characterise interesting events in ten years of data taking, starting in 2026.
Look up the CODE
