The 2H(α, γ)6Li experiment at LUNA

Recent observations of 6Li in metal poor stars suggests a huge production of this isotope during BBN, through the 2H(α, γ)6Li reaction. This reaction has never been directly measured inside the BBN energy region because its cross section drops steeply at low energy, while indirect measurements using the Coulomb dissociation of 6Li only give upper limits due to the dominance of nuclear breakup process. Exploiting the ultralow background at the 400 keV LUNA accelerator, located deep underground in Italy’s Gran Sasso laboratory, for the first time the reaction has been measured directly at BBN energy. The LUNA data and their implications for the BBN theory are discussed.


Introduction
In its standard picture, the Big Bang Nucleosynthesis occurs during the first minutes of universe, with the formation of light isotopes such as D, 3 He, 4 He, 6 Li and 7 Li.Their abundance only depends on standard model physics, on the baryon-to-photon ratio η and on the nuclear cross sections of involved processes.The observations of D, 3 He, and 4 He abundances are in good agreement with calculations, confirming the overall validity of BBN theory.On the other hand, the observed abundance of 7 Li is a factor 2-4 lower than the predicted one, while the amount of 6 Li observed in metal poor stars is unexpectedly large compared to Big Bang Nucleosynthesis (BBN) predictions, about 3 orders of magnitude higher than the calculated value [1].Even though many of the claimed 6 Li detections are controversial, for a few metal-poor stars there still seems to be a significant amount of 6 Li [1].The difference between observed and calculated abundances of 7 Li and 6 Li may reflect unknown post-primordial processes or physics beyond the Standard Model, and are reported in literature as the "Primordial Lithium problems" [1,2].In standard BBN, the leading process to synthesize 6 Li is the 2 H(α, γ) 6 Li reaction.The 2 H(α, γ) 6 Li cross section is very small at BBN energies (30 E(keV) 400), because of the Coulomb barrier and because electric dipole transition is suppressed for the iso-scalar particles 2 H and 4 He [3].Therefore, the 2 H(α, γ) 6 Li cross section has been measured only for energies greater than 1 MeV and around the 711 keV resonance [4,5].There are two indirect attempts to determine the 2 H(α, γ) 6 Li cross section at BBN energies, using the Coulomb dissociation technique [6,7].In this approach, the time-reversed reaction 6 Li(γ, α) 2 H is studied using an energetic 6 Li beam and a target of high nuclear charge.However, it is difficult to unfold the cross section with this method, because the nuclear effects are dominant and the result strongly depends on the theoretical assumptions [3,7].The present work reports on the first direct measurement of the 2 H(α, γ) 6 Li at BBN energies, performed by a e-mail: carlo.gustavino@roma1.infn.it the LUNA collaboration (LUNA=Laboratory for Underground Nuclear Astrophysics).The measurement has been carried with the world only underground accelerator, situated at the LNGS laboratory (LNGS=Laboratorio Nazionale del Gran Sasso), Italy [8].The "Gran Sasso" mountain provides a natural shielding which reduces the muon and neutron fluxes by a factor 10 6 and 10 3 , respectively.The suppression of the cosmic ray induced background also allows an effective suppression of the γ-ray activity by a factor 10 2 -10 5 , depending on the photon energy [9].The ultra-low background at LNGS made possible to study the 2 H(α, γ) 6 Li reaction at Big Bang energies.In the following, the new data and their implications for Big Bang nucleosynthesis will be shown.Fig. 1 shows the experimental set-up used for the 2 H(α, γ) 6 Li reaction.The measurement is based on the use of the 400 kV accelerator, that provides an α-beam of high intensity.The α-beam impinges a windowless gas target of D 2 , with a nominal operating pressure of 0.3 mbar.The signal is maximized by stretching the beam intensity up to about 350 µA and by using a configuration with the high purity Ge(Li) detector (HPGe) as close as possible to the beam line.The natural background of LNGS is further reduced by means of a 4π lead shield around the reaction chamber and the HPGe detector.The set-up is enclosed in a anti-radon box flushed with high purity N 2 , to reduce and stabilize the γ activity due to the radon decay chain.The measurement of the 2 H(α, γ) 6 Li reaction is affected by an inevitable beam induced background.In fact, the 2 H(α, α) 2 H Rutherford scattering induces a small amount of 2 H( 2 H, n) 3 He (Q = 3.267 MeV) and 2 H( 2 H, p) 3 H (Q = 4.033 MeV) reactions.While the 2 H( 2 H, p) 3 H reaction is not a problem in this context, the neutrons produced by the 2 H( 2 H, n) 3 He reaction induce (n, n γ) reactions in the HPGe detector and in the surrounding materials (lead, steel, copper), generating a neutron induced background (here and after NIB) in the HPGe spectrum.To reduce the effective path for the scattered deuterons, and therefore the yield of the 2 H( 2 H, n) 3 He reaction, a removable 177 mm long tube, with a square cross section of 17 × 17 mm, is placed along the beam line (see Fig. 1).In this way, the neutron production is limited at the level of few neutrons/second.The set-up is implemented with a silicon detector faced to the gas target volume, to monitor the running conditions through the detection of protons generated in the 2 H( 2 H, p) 3 H reaction (E p 3 MeV).As a matter of fact, the proton rate is related to the number of produced neutrons, since the cross sections of the two conjugate 2 H( 2 H, n) 3 He and 2 H( 2 H, p) 3 H reactions are similar and well known [10].Further details on the set-up can be found in [11].

Measurement
The HPGe spectrum has three components: the natural background (BCK), the neutron induced background (NIB) and the signal due to the γ-rays produced in the 2 H(α, γ) 6 Li reaction.To extract the signal due to the 2 H(α, γ) 6 Li reaction it is therefore necessary to subtract BCK and NIB.The BCK is stable and low due to the lead shielding and the anti-radon box.Therefore, it has been measured separately in long off-beam measurements and properly subtracted.The NIB spectrum at LUNA EPJ Web of Conferences 07009-p.2 has been extensively studied by means of experimental data and a detailed simulation based on the GEANT4 package [11].Fig. 2 shows the energy distribution of neutrons at E beam = 400, 280 keV.It is worth to point out that the neutron energy distribution depends weakly on the beam energy, because the difference between the two beam energies ∆E beam = 120 keV is much smaller with respect to energy of neutrons produced in the 2 H( 2 H, n) 3 He reaction (E n = 2450 keV in the center-of-mass system).As a consequence, the NIB shape is almost unaffected while changing the α-beam energy, because it is generated by the neutron interaction with the material surrounding the HPGe detector [11].The energy of γ-rays produced by the 2 H(α, γ) 6 Li reaction (Q = 1473.48keV) depends on kinematics through the following relationship: In this equation, m D and m α are the masses of deuteron and α particles, respectively.In our set-up The Doppler correction is ∆E doppler ≈ 16 keV at E α = 400 keV, and the γ-ray energy shift due to the recoiling compound nucleus is about E Recoil = 0.2 keV.As the γ-rays produced in the 2 H(α, γ) 6 Li reaction depends on the beam energy, the NIB can be subtracted performing in-beam measurements at two different energies, that is: 1. Measurements with E beam = 400 keV on D 2 target.The Ge spectrum is mainly due to the background induced by neutrons.The 2 H(α, γ) 6 Li γ signal is kinematically constrained in a well defined region of interest (RoI), 1587 < E γ (keV) < 1625.
2. Same as 1., but with E beam = 280 keV.In this case, the RoI of γ-rays produced in the 2 H(α, γ) 6 Li reaction is shifted to the 1550 − 1580 keV region.
As the RoIs of these two measurements do not overlap, the signal at 400 (280) keV can be extracted by subtracting the NIB obtained in the 280 (400) keV measurement.The in-beam measurements have been performed by alternating the two energies during the ∼ 40 days of acquisition time (about 20 days for each beam energy).It has been found that the NIB rate is not related in a simple way to the working parameters such as the beam energy, beam intensity and target density.In fact, it slightly increases with time because of the progressive implantation of scattered deuterium in the material surrounding the beam line, such as the 177 mm long tube coaxial to the beam line (see fig. 1).As a consequence, the scattered deuterons interact not only with the deuterium gas target but also with the implanted deuterons, making the NIB rate dependent on the operation time.Therefore, the NIB 280 has been normalized to the NIB 400 one with a minimization procedure, leaving the normalization factor as a free parameter.Fig. 3 shows the spectra obtained with nominal beam energies of E α = 400, 280 keV, respectively (BCK subtracted).The spectrum at E α = 280 keV have been normalized accordingly with the minimization procedure.The counting excess due to the 2 H(α, γ) 6 Li reaction is clearly visible in the E α = 400 keV RoI.The shape of the counting excess suggests a forward-backward asymmetry of emitted photons, possibly due to the interference between dipole and quadrupole transitions [3].The reaction yield at E α = 280 keV is slightly lower with respect the one obtained at E α = 400 keV, as a consequence of the higher Coulomb barrier and the absence of resonant nuclear effects.In the low energy domain, the cross section σ(E) can be expressed using the astrophysical factor S (E), defined by the formula: S (E) contains all the nuclear effects and, for non-resonant reactions, it is a smoothly varying function of energy.The exponential term takes into account the Coulomb barrier.The Sommerfeld parameter η is given by 2πη = 31.29Z 1 Z 2 (µ/E) 1/2 .Z 1 and Z 2 are the nuclear charges of the interacting nuclei.µ is their reduced mass (in units of a.m.u.), and E is the center of mass energy (in units of keV).Fig. 4 shows the preliminar S 24 astrophysical factor obtained by LUNA as a function of the centerof-mass energy, together with all previous direct measurements [4,5].As shown in the figure, the S 24 astrophysical factor measured by LUNA is slightly in agreement with the theoretical S 24 factors describing the Coulomb dissociation measurement by [7].The astrophysical factor of the 2 H(α, γ) 6 Li reaction as a function of the center-of-mass energy.The preliminar result of the LUNA measurement (statistical error only) is shown with all the previous direct [4,5] measurements.The continuous lines show the theoretical E1, E2, and total S 24 factors describing the Coulomb dissociation measurement by [7].The BBN energy of interest (orange band) is also shown.

DOI: 10
.1051/ C Owned by the authors, published by EDP Sciences,

Figure 3 .
Figure 3. (Colour online) Experimental spectra for E α = 400 keV (black line) and for E α = 280 keV (red line).The natural background has been subtracted.The narrow peaks at 1547 keV and 1623.5 keV are due to the de-excitation of 63 Cu and 65 Cu isotopes induced by inelastic neutron scattering [11].The red and violet bands indicate the RoI at E α = 400 keV RoI and E α = 280 keV, respectively.Note the counting excesses visible (green bins) in correspondence to the RoIs.

Figure 4 .
Figure 4. (Colour online) The astrophysical factor of the 2 H(α, γ)6 Li reaction as a function of the center-of-mass energy.The preliminar result of the LUNA measurement (statistical error only) is shown with all the previous direct[4,5] measurements.The continuous lines show the theoretical E1, E2, and total S 24 factors describing the Coulomb dissociation measurement by[7].The BBN energy of interest (orange band) is also shown.