GALLEX

The surface buildings of the Gran Sasso laboratory near Assergi. The GALLEX gallium tank stood in the halls deep under the massif.
The surface buildings of the Gran Sasso laboratory near Assergi. The GALLEX gallium tank stood in the halls deep under the massif.Photo: TQB1, CC BY-SA 3.0, Wikimedia Commons

GALLEX — the gallium experiment — was a radiochemical detector for solar neutrinos at the Gran Sasso underground laboratory. It ran from 1991 to 1997 and was subsequently continued as GNO until 2003.

Its significance lies in one number: 233 kiloelectronvolts.

Why gallium

The Homestake experiment worked with chlorine and could therefore only detect neutrinos above 814 keV. With that it captured only the rare high-energy neutrinos from boron-8 decay — less than a hundredth of a per mille of all solar neutrinos.

The Sun's main reaction, the pp chain, delivers neutrinos of at most 420 keV. They remained invisible to chlorine.

Gallium changes that. The transition from gallium-71 to germanium-71 has a threshold of only 233 keV. With that, for the first time, the portion originating from the fundamental reaction became accessible.

And that is more than a technical improvement. The rate of pp neutrinos can be predicted almost without model assumptions: it follows directly from the luminosity of the Sun, since every unit of energy generated is accompanied by a certain number of neutrinos. Whoever measures this rate is not testing a solar model but an almost inescapable balance.

The set-up

GALLEX used around 30 tonnes of gallium, dissolved as gallium chloride in hydrochloric acid — in total about a hundred tonnes of solution in a tank beneath 1,400 metres of rock.

Every three to four weeks the germanium that had formed was flushed out. This concerned a few dozen atoms in a hundred tonnes of solution.

The procedure: a gas stream carries out the volatile germanium tetrachloride, it is concentrated in several stages, converted into germane and filled into tiny counter tubes. There it decays over months, and every decay is registered individually.

The half-life of germanium-71 is 11.4 days — short enough to count almost everything within a few months, long enough to survive the extraction.

The calibration with an artificial sun

One objection suggested itself: how does one know that the chemistry really finds every atom and that the counter tubes work correctly?

GALLEX answered that in the most convincing way possible. The collaboration had an artificial neutrino source made from chromium-51 — chromium specially irradiated in a reactor, whose activity was precisely known. This source was lowered into the tank, and it was checked whether the measured rate corresponded to the known source strength.

Two such calibrations were carried out, in 1994 and 1995. The procedure worked — a positive control of a kind rarely possible so directly in physics.

The result

GALLEX found about 60 per cent of the expected rate. The solar neutrino problem therefore existed at low energies too, but considerably more weakly than at Homestake, where only a third arrived.

This pattern was the actual find. An error in the solar model could hardly have acted in this energy-dependent way. A transformation along the way, by contrast, could — and precisely in the manner predicted by the MSW effect.

Together with SAGE, GALLEX thus supplied a strong argument years before SNO settled the matter.

The anomaly

In the calibration measurements something unexpected showed up: the measured rate lay slightly below the one calculated from the source strength. SAGE found the same.

This gallium anomaly has not been explained to this day. It is discussed as a possible indication of sterile neutrinos — or as a sign that the underlying nuclear transition is less well known than assumed.

Sources

  • GALLEX Collaboration: Solar neutrinos observed by GALLEX at Gran Sasso, Physics Letters B 285, 376 (1992).
  • GALLEX Collaboration: First results from the ⁵¹Cr neutrino source experiment with the GALLEX detector, Physics Letters B 342, 440 (1995).