Thierry Stolarczyk

Cosmic particle hunter

GALLEX
Gallex (GALLium EXperiment) was a pionneering experiment aiming at detecting solar neutrinos, in particular the primordial solar neutrinos produced the core of the Sun during the very first fusion reactions of hydrogen into helium. In 1992 and during the following years, GALLEX contributed to show that the solar neutrino deficit, as measured since 1970, was not due to an incomplete understanding of the solar fusion processes but to the transformation of neutrinos during their travel to the Earth.

This video clip was made in 1992 for a scientific outreach broadcast that was cancelled (the clip was never shown) - (c) CEA, 1992

 

The text below is extracted, adapted and translated from my PhD thesis document "Bruits de fond dans l'expérience GALLEX de détection des neutrinos solaires" (introduction)

The neutrino flux emitted during the fusion of hydrogen in the core of the Sun is a direct probe of the solar interior processes since, contrarily to photons, neutrinos escape the stellar matter essentially without interactions. Their detection allows probing the Solar Standard Model that derives from observable characteristics (radius, luminosity...) the Sun internal properties (temperature distribution, matter density...) and the sites where the solar neutrinos are produced, and how abundant they are, depending on the nuclear reaction they originate from.

The most numerous neutrinos, more than 90% of the total, come from the primordial fusion reaction of two protons (p + p → 2H + e+ + νe). They have also the advantage to have a flux which is easily predictible by the standard model.

The GALLEX radiochemical experiment chose to detect these neutrinos thanks to the νe + 71Ga → 71Ge + e- reaction (threshold: 233 keV) where the interaction of the neutrino is signed by the observation of the 71Ge decay. GALLEX used 30 tons of natural gallium under the form of a solution of gallium chloride. It was installed in the Gran Sasso mountainous massif in Italy, under 1500 metres of rock.

GALLEX was not only devoted to the detection of the primordial neutrinos, it also contributed to solve the so-called "solar neutrino problem" that put the solar standard model into difficulties and triggered fascinating new ideas in particle physics. This problem was definitively solved in 2011. It was born with the chlorine experiment in the early 70's that measured one third of the predicted flux for decennia.

This experiment was using the reaction 37Cl + νe → 37Ar + e- (threshold : 814 keV), the clhorine being under the form of 400000 litres of C2Cl4, a liquid stain remover. It was sensitive essentially to boron neutrinos (8B → 8Be* + e+ νe), a source with a much higher mean energy, a more uncertain prediction, and a flux 10000 times lower than the primordial neutrinos.

The disagreement was confirmed in 1989 by the Kamiokande II experiment. It used a completely different way to detect the neutrinos of originating from the same reaction, observing the Cherenkov light prodiuced by the electrons emitted by the diffusion of neutrinos in water (The successor of Kamiokande, SuperKamiokande, is still in operation).

The number of interactions due to neutrinos is extremelly small. In the chlorine experiment there was only 0.5 37Ar produced per day in the target tank. In GALLEX, 1,2 71Ge were expected to be produced per day in the 54 m3 of gallium solution. Because the neutrino-electrons (νe) from the Sun disappear during their travel to Earth, less than an atom was produced per day.

The detector and the collaboration

The institutes in the GALLEX collaboration in 1990 (A that time Germany was not reunified yet. FRG stands for Federal Republic of Germany, the former West Germany):
  • Max-Planck Institut für Kernphysik (Heidelberg-FRG)
  • Kernforschungszentrum Karlsruhe - KFK (FRG)
  • Dip. di fisica dell'Universita Milano (Italy)
  • Laboratori Nazionali del Gran Sasso - INFN (Italy)
  • Physik Dept.E15 - Technische Universität München (FRG),
  • Université de Nice-Observatoire (France)
  • Universita di Roma II (Italy)
  • Centre d'Etudes Nucléaires de Saclay (CEA Saclay, France)
  • Weizmann Institute of sciences - Rehovot (Israel)
  • Brookhaven National Laboratory (USA).
  • The first gallium experiment was proposed in 1978. In the years 1979-1982, a pilot experiment was conducted by teams from the Brookhaven National Laboratory and the Max-Planck Institut of Heidelberg. It used 1.26 tons of gallium in the form of 4.26 tons of a GaCl3 + HCl solution. This experiment showed that the detection of solar neutrinos was possible provided that 30 tons of gallium are made available.

    In 1985 the collaboration was formed with 50 scientists and technicians, from 10 institutes (See boxed text).

    The detector was placed in the Gran Sasso underground laboratory, near the city of l'Aquila, 125 km North-East of Roma. This laboratory is situated 6.3 km from the motorway tunnel entrance under 1500 m of rock (equivalent to 3500 m of water) reducing the background cosmic ray flux by a factor 100 000. The underground laboratory comprises three exeprimental halls. The first one, hall A, hosted the Gallex Main Building and its associated Low Level Building.

    The main building

    The Gallex Main Building hosted the target tank (Tank A). Its total volume was 70 m3 and it was filled with 30 tons of natural gallium in the form of an extremely corrosive liquid solution of  GaCl3 + HCl (volume : 54 m3). Natural gallium is composed of 39.9% of 71Ga and 60.1% of 69Ga.

    A second tank (Tank B) was close by. It was quasi identical to the later and was planned to transfer the gallium solution during the maintenance of tank A, or in case of an incident. Both tanks were in polyester “PALATAL” reinforced with glass fibers and were internally coated with PVDF (teflon) to resist to the solution acidity. Moreover their manufacturing was subject to drastic constraints in order to reduce the intrinsic radioactive contamination.

    This building also hosted the experimetal setup for the germanium extraction and the chemical bench used to transform it into GeH4. This gaz was then pushed into a small proportionnal counter which was then transported to the Low Level Building where the 71Ge decays could be observed.

    The counting building

    Map of the experimental halls in the Gran Sasso underground laboratory (as in 1990). GALLEX and the MACRO experiment (search for magnetic monopoles) are indicated

    This building called Low Level Building hosted the counting and data acquisition. On the first floor there was a Faraday cage, a cylindrical counting chamber isolating the active counters from the exteranl infleunces. These counters were also protected with shieldings either active (cylindrical Soidum Iodide scintillator) or passive (lead, iron, low radioactivity copper) against the rare cosmic ray muons reaching this depth. The data from the counters, once preamplifies, were transmitted to the upper floor with a fiber optic connection. The data acquision was performed using two coupled computer (microVax).

    According to the standard model, the daily 71Ge rate from solar neutrinos was 1,2. Since the 71Ge decay lifetime is 11.43 days, every three weeks a dozen of atoms were extracted.

    The germanium in the solution, either the radioactive one produced by solar neutrinos and the background processes, or the macroscopic amount (about 1 mg) of natural non radioactive isotopes introduced at the start of each exposure, bind with the chlorine atoms to form GeCl4 molecules. This compound is extremely volatile in an acid medium. It extraction was easiy realised by flushing several thousands of litres of dry nitrogen into the solution.

    The nitrogen was then put in contact with pure water in which GeCl4 easily dissolves. Then the GeCl4 was collected and transformed into germane (GeH4). Xenon is then added to this gas in order to obtain a  70% Xe - 30% GeH4 mixture which has genuine characteristics for the counting (energy resolution, electron drift veolcity, background rejection). The gas mixture was then pushed into a proportionnal counter.

     

    Picture of tank A, of the GeCl4 exchanger, and a drawing showing the full extraction system (a calibration neutrino source used in 1995 is shown in the center of the tank).

    The proportionnal counters

    Miniaturised proportionnal counter used in GALLEX (here from the latest generation in 1990, called “HD-II”, lthe silver cylinder on the left is the counting volume, it is 25 mm long for a volume of about 1 cm3)

    The proportionnal counters inherit from those build for the chlorine experiment. They are made of quartz. Their 1 μm thick cathode is obtained from silicon or iron evaporation on the internal part of the counting volume, whereas their anode was a 7.5 μm diametre tungsten wire. The electron capture decays of 71Ge atomes are signed by a K-peak at 10.4 keV and a L-peak at 1.2 keV (Auger electrons or X-rays). The counter intrinsic background has been remarkably reduced using very low radioactivity materials. Each signal in the counter is characterised by the deposited energy, its time and the signal rise-time, the latter being a very powerfull discriminator of the parasitic pulses. The total background daily rate in the counters does not excedd 0.15 inthe L-peak and 0.02 in the L-peak for the best of them.