Study of the 2H(α,γ)6Li reaction at LUNA

(PhD project of Michael Anders, supported by DFG BE 4100/2-1)

Astrophysical background and motivation:

According to the standard Big Bang model, a few minutes after Big Bang the first chemical elements were formed: Apart from the already existing protons (1H) especially helium (4He with traces of 3He), some deuterium (2H), and small amounts of lithium, about 10-10 with respect to hydrogen. This primordial lithium consists of the stable isotopes 7Li (mainly) and some 6Li. The most important primordial production process for 6Li is the 2H(α,γ)6Li reaction (Serpico et al. 2004).

A few minutes after the Big Bang, primordial nucleosynthesis stopped. Over the remainder of the history of the universe, the isotopes of lithium (6Li, 7Li), beryllium (9Be, 10Be) and boron (10B, 11B) are synthesized by spallation reactions of energetic cosmic rays with the interstellar medium. These nuclei are too fragile to be produced in stars in significant amounts; instead, it is believed that they are even depleted there. Conventional wisdom holds that while primordial processes are responsible for most 7Li, cosmic ray spallation produces most 6Li.

In the early 1980s Spite and Spite found a constant amount of lithium in stars of a wide range of metallicity (this is the amount of the elements except hydrogen and helium, compared to the amount of hydrogen). Even very old stars having a metallicity 103 or 104 times less than our sun show a constant lithium abundance. This leads to the conclusion that the lithium abundance found in these stars is not synthesized over the history of the universe, but mostly primordial, confirming the model.

However, new data from the analysis of the tiny isotopic shift in lithium absorption lines seem to indicate a similar plateau also for the less abundant isotope 6Li (Asplund et al. 2006). Even for very old stars, the abundance ratio 6Li/7Li is consistent with the value of 0.07 observed in the solar system. This observation is in contrast with the presently assumed model that 6Li is produced over time by cosmic ray spallation. It could be explained with a very high primordial 6Li production, about 100 times higher than present models of big bang nucleosynthesis. Such an increase could be caused either by a higher rate of the  2H(α,γ)6Li nuclear reaction, or by new physics (see e.g. Jedamzik et al. 2004).

Measuring the 2H(α,γ)6Li reaction cross sections is a very exciting problem in this case. The reaction has never been measured in the big bang nucleosynthesis energy window before. Previous coulomb dissociation data are questionable due to a dominant non-coulomb (nuclear) breakup component, therefore a direct measurement is necessary.

Our approach:

The LUNA2 accelerator provides ion beams (1H+, 4He+ and potentially 3He+) with an energy of up to 400 keV. It's located at the Gran Sasso underground laboratory, about 1400 m of rock provide an excellent shield against cosmic radiation. Because of the very low expected cross section to be studied, the apparatus must be able to distinguish some tens of events per day from the background. Apart from the advantages of the location, a lead castle in a radon box reduces the background from natural radionuclides.

The photons are detected by a high-purity germanium detector with a relative efficiency of 137%. Its end cap is less than 10 cm away from the target. The target consists of a differentially pumped windowless gas target chamber containing the deuterium (2H) gas at a standard operating pressure of 0.1-1.0 mbar. An array of roots and turbomolecular pumps between the gas target and the accelerator keeps its vacuum. The alpha (4He+) beam is finally stopped at a beam calorimeter where the beam intensity is measured.

The LUNA d+alpha setup before mounting the lead castle

The 2H(α,γ)6Li setup at LUNA before mounting the lead castle.

However, there is a problem to solve or at least to control: The 4He+ beam accelerates deuterons through elastic scattering. They can interact with other deuterons or be implanted at metal surfaces (e.g. the calorimeter end cap), being potential targets for other deuterons also accelerated by scattering. Through the d(d,n)3He reaction, a tiny neutron flux is induced. This flux doesn't induce any safety hazard, but disturbs the experiment anyway: The neutrons produce photons through inelastic scattering, unfortunately also in the γ-ray energy region of interest (at 1.6 MeV). This beam induced background has to be known and reduced to a minimum.

Fortunately the reaction d(d,p)t has a similar cross section as d(d,n)3He. One silicon detector measures the proton flux to determine the neutron flux indirectly.

Simulations have shown that a similar neutron energy spectrum can be achieved also with a lower beam energy. In this case the beam induced background remains the same, but the  2H(α,γ)6Li signal is shifted to lower gamma energies. Subtracting such spectra from each other leaves two  2H(α,γ)6Li signals.

After understanding and improving the setup within several years, the first phase of data acquisition was finished on end of October 2010, the second one in April 2011. With the final setup and final measurement parameters, long-time measurements were performed in May, June, October and November 2011. Additional experiments in 2012 to further study the neutron interaction in HPGe detectors completed the data set. The data analysis has been completed, the publication of the results is in preparation.

See also the recent article which describes the setup and discusses the observed beam induced background:

M. Anders, D. Trezzi, A. Bellini et al.: "Neutron induced background by an α-beam incident on a deuterium gas target and its implications for the study of the 2H(α,γ)6Li reaction at LUNA", European Physical Journal A, 49:28 (2013)