The uncertainties which still plague our understanding of the evolution of the light nuclides D, 3He and 4He in the Galaxy are described. Measurements of the local abundance of deuterium range over a factor of 3. The observed dispersion can be reconciled with the predictions on deuterium evolution from standard Galactic chemical evolution models, if the true local abundance of deuterium proves to be high, but not too high, and lower observed values are due to depletion onto dust grains. The nearly constancy of the 3He abundance with both time and position within the Galaxy implies a negligible production of this element in stars, at variance with predictions from standard stellar models which, however, do agree with the (few) measurements of 3He in planetary nebulae. Thermohaline mixing, inhibited by magnetic fields in a small fraction of low-mass stars, could in principle explain the complexity of the overall scenario. However, complete grids of stellar yields taking this mechanism into account are not available for use in chemical evolution models yet. Much effort has been devoted to unravel the origin of the extreme helium-rich stars which seem to inhabit the most massive Galactic globular clusters. Yet, the issue of 4He evolution is far from being fully settled even in the disc of the Milky Way.

Thermohaline mixing has been recently identified as the dominating process that governs the photospheric composition of low-mass bright giant stars (Charbonnel & Zahn 2007a). Here we present the predictions of stellar models computed with the code STAREVOL that takes into account this mechanism together with rotational mixing and atomic diffusion. We compare our theorical predictions with recent observations and discuss how the corresponding yields for 3He are compatible with the observed behaviour of this light element in our Galaxy.

Population III stars initiated the chemical enrichment of the Universe. Chemical evolution models seem to favour fast rotators among the very low-metallicity population. When a star rotates fast, it ejects significant quantities of He and its nucleosynthesic products are modified compared to the case without rotation. The value of ΔY/ΔO is explored from a theoretical point of view through stellar models of zero- or very low-metallicity.

Lithium represents a key element in cosmology, as it is one of the few nuclei synthesized during the Big Bang. The primordial abundance of 7Li allows us to impose constraints on the primordial nucleosynthesis and on the baryon density of the universe. However, 7Li is not only produced during the Big Bang but also during galactic evolution: measures of stellar Li in our Galaxy suggest an almost constant Li abundance (the so-called Spite plateau) at low metallicities and a subsequent increase in the disk stars, leading to a Li abundance in Population I stars higher by a factor of ten than in Population II stars. This means that there must exist several possible stellar sources of 7Li: asymptotic giant branch stars, supernovae, novae, red giant stars. 7Li is also partly produced in spallation processes while 6Li is entirely produced by such processes. All of these sources have been included in galactic chemical evolution models and constraints have been derived on the primordial 7Li and its evolution, as well on stellar models. I will review these models and their results and what we have learned about 7Li evolution. Some still open problems, such as the disagreement between the primordial 7Li abundance as derived by WMAP and as measured in Population II stars, and the uncertainties about the main sources of stellar 7Li will be discussed.

Type Ic supernova (SN Ic) is the gravitational collapse of a massive star without H and He layers. It propels several solar masses of material to the typical velocity of 10,000 km/s, a very small fraction of the ejecta nearly to the speed of light. We investigate SNe Ic as production sites for the light elements Li, Be, and B, via the neutrino-process and spallations. As massive stars collapse, neutrinos are emitted in large numbers from the central remnants. Some of the neutrinos interact with nuclei in the exploding materials and mainly 7Li and 11B are produced. Subsequently, the ejected materials with very high energy impinge on the interstellar/circumstellar matter and spall into light elements. We find that the ν-process in the current SN Ic model produces a significant amount of 11B, consistent with observations if combined with B isotopes from the following spallation production.

My aim is to show how the abundance ratios of the light elements (6 to 11) are related to the properties of the strong nuclear interaction and, in particular, to the major influence of closed shells of neutrons and protons, (the magic numbers : 2, 8, etc) on the binding energies of the nuclei.

A short overview is presented of current issues concerning the production and evolution of Li, Be and B in the Milky Way. In particular, the observed “primary-like” evolution of Be is re-assessed in the light of a novel idea: it is argued that Galactic Cosmic Rays are accelerated from the wind material of rotating massive stars, hit by the forward shock of the subsequent supernova explosions. The pre-galactic levels of both Li isotopes remain controversial at present, making it difficult to predict their Galactic evolution. A quantitative estimate is provided of the contributions of various candidate sources to the solar abundance of Li.

The single stable isotope of beryllium is a pure product of cosmic-ray spallation in the ISM. Assuming that the cosmic-rays are globally transported across the Galaxy, the beryllium production should be a widespread process and its abundance should be roughly homogeneous in the early-Galaxy at a given time. Thus, it could be useful as a tracer of time. In an investigation of the use of Be as a cosmochronometer and of its evolution in the Galaxy, we found evidence that in a log(Be/H) vs. [α/Fe] diagram the halo stars separate into two components. One is consistent with predictions of evolutionary models while the other is chemically indistinguishable from the thick-disk stars. This is interpreted as a difference in the star formation history of the two components and suggests that the local halo is not a single uniform population where a clear age-metallicity relation can be defined. We also found evidence that the star formation rate was lower in the outer regions of the thick disk, pointing towards an inside-out formation.

I am pleased to recall that the first determination of the Li–abundance in the Sun was made at the Geneva Observatory in 1975 by Edith Müller, Eric Peytremann and Ramiro de la Reza on the basis of spectra taken at Kitt Peak. The first Be determination was also made in Geneva the same year by Y. Chmielewski, J. Brault (Kitt Peak National Observatory) and E. Müller. These two outstanding works opened the door for all further investigations on these light elements.

In Discussion D the following problems were addressed: Has 6Li really been detected in the atmospheres of metal-poor halo stars? Is there a downward trend or increased scatter of Li abundances in stars on the ‘Li-plateau’ at metallicities [Fe/H] ≲ −2.5? Are there significant differences of Li abundances in main-sequence, turn-off, and sub-giant stars in globular clusters? Is the Li abundance in solar-type stars related to the presence of planets? How does the Be abundance in dwarf stars increase with the heavy-element abundance, and is there a cosmic scatter in Be at a given [Fe/H]? The discussion of these problems is summarized and some suggestions for future observational and theoretical studies are mentioned.

Chemical features of the local disk have firmly established the picture for the formation of the Galactic disk that the star formation has proceeded under the continuous accretion of low-metallicity gas from the halo. It sets two determinant processes for the evolution of deuterium (D), that is, the destruction of D in the interior of stars and the supply of new (nearly) primordial D associated with the gas infall. Conventional Galactic chemical evolution (GCE) models predict that this scheme leads to a monotonic decrease in D/H with time and ends up in the present-day D/H abundance (D/H)0 which is severely lower than the recently observed estimates. These predicted features are the natural results of a construction of the metal-rich (~solar abundance) local star+gas system. Here we propose that the new GCE models, that incorporate large-scale winds form the Galactic bulge which entrain heavy elements and drop them on the disk with the recent tendency of star formation in tune with the observed implications, make the system rich in both metals and D. In addition, our finding of a gradual increase in D/H with time during the last several Gyr is observationally supported by the D/H abundance for the protosolar cloud lower than (D/H)0.