Big bang nucleosynthesis updated with the NACRE compilation

We update the Big Bang Nucleosynthesis calculations on the basis of the rec ent NACRE compilation of reaction rates. The average values of the calculat ed abundances of light nuclei do not differ significantly from those obtain ed using the previous Fowler's compilation. However, Li-7 is slightly depre ssed at high baryon to photon ratio eta. Concerning B-10, its abundance is significantly lower than the one calculated with the Caughlan & Fowler (198 8) rates as anticipated by Rauscher & Raimann (1997). We estimate the uncer tainties related to the nuclear reaction rates on the abundances of D, He-3 , He-4, Li-6, Li-7, Be-9, B-10 and B-11 of cosmological and astrophysical i nterest. The main uncertainty concerns the D(p, gamma)He-3 reaction rate af fecting the synthesis of Li-7 at rather high baryonic density and also the He-3(alpha, gamma)Be-7 and Li-7(p,alpha)He-4 reactions. On the left part of the lithium valley the uncertainty is reduced due to the improvement of th e measurement of the T(alpha, gamma)Li-7 reaction rate. The observed abunda nces of the nuclei of interest are compared to the predictions of the BBN m odel, taking into account both observational and theoretical uncertainties. Indeed, the Li-7 abundance observed in halo stars (Spite plateau) is now d etermined with high precision since the thickness of this plateau appears, in the light of recent observations, exceptionnaly small (< 0.05 dex). The potential destruction/dilution of Li-7 in the outer layers of halo stars wh ich could mask the true value of the primordial abundance is in full debate , but the present trend is towards a drastic reduction of the depletion fac tor (about 0.10 dex). It is why we use this isotope as a preferred baryomet er. Even though much efforts have been devoted to the determination of deut erium in absorbing clouds in the line of sight of remote quasars, the stati stics is very poor compared to the long series of lithium measurements. Tak ing into account these lithium constraints, two possible baryonic density r anges emerge, eta(10) = 1 5-1.9 and eta(10) = 3.3-5.1. In the first case, L i is in concordance with D from Webb et al. (1997) and 4He from Fields & Ol ive (1998) and Peimbert & Peimbert (2000). In the second case, agreement is achieved with D from Tytler et al. (2000) and He-4 from Izotov & Thuan (19 98). Concerning the less abundant light isotopes, the theoretical BBN abundance of Li-6 is affected by a large uncertainty due to the poor knowledge of the D(alpha, gamma)Li-6 reaction rate. However, at high eta, its abundance is so low that there is little chance to determine observationally the true BB N Li-6 abundance. But, at low eta, its abundance being one thousandth of th at of primordial Li-7, 6/7 ratio measurements at very low metallicity are n ot totally hopeless in the future. Nevertheless, in the present situation, Li-6 is cosmologically relevant, though indirectly, since its mere presence in a few halo stars, corroborates the fact that it is essentially intact i n these stars together with Li-7 and thus the Spite plateau can be used as such to infer the primordial Li-7 abundance. The Be and B abundances produc ed in the Big Bang are orders of magnitudes lower, and spallation of fast c arbon and oxygen is probably their unique source, in the early Galaxy.