Published in Physica Scripta, 85, p. 12, 2000


David Tytler, John M. O'Meara, Nao Suzuki & Dan Lubin

Center for Astrophysics and Space Sciences
University of California, San Diego
MS 0424; La Jolla; CA 92093-0424

Abstract. Big Bang Nucleosynthesis (BBN) is the synthesis of the light nuclei, Deuterium, 3He, 4He and 7Li during the first few minutes of the universe. This review concentrates on recent improvements in the measurement of the primordial (after BBN, and prior to modification) abundances of these nuclei. We mention improvement in the standard theory, and the non-standard extensions which are limited by the data.

We have achieved an order of magnitude improvement in the precision of the measurement of primordial D/H, using the HIRES spectrograph on the W. M. Keck telescope to measure D in gas with very nearly primordial abundances towards quasars. From 1994 - 1996, it appeared that there could be a factor of ten range in primordial D/H, but today four examples of low D are secure. High D/H should be much easier to detect, and since there are no convincing examples, it must be extremely rare or non-existent. All data are consistent with a single low value for D/H, and the examples which are consistent with high D/H are readily interpreted as H contamination near the position of D.

The new D/H measurements give the most accurate value for the baryon to photon ratio, eta , and hence the cosmological baryon density. A similar density is required to explain the amount of Lyalpha absorption from neutral Hydrogen in the intergalactic medium (IGM) at redshift z appeq 3, and to explain the fraction of baryons in local clusters of galaxies.

The D/H measurements lead to predictions for the abundances of the other light nuclei, which generally agree with measurements.

The remaining differences with some measurements can be explained by a combination of measurement and analysis errors or changes in the abundances after BBN. The measurements do not require physics beyond the standard BBN model. Instead, the agreement between the abundances is used to limit the non-standard physics.

New measurements are giving improved understanding of the difficulties in estimating the abundances of all the light nuclei, but unfortunately in most cases we are not yet seeing much improvement in the accuracy of the primordial abundances. Since we are now interested in the highest accuracy and reliability for all nuclei, the few objects with the most extensive observations give by far the most convincing results.

Earlier measurements of 4He may have obtained too low a value because the He emission line strengths were reduced by undetected stellar absorption lines. The systematic errors associated with the 4He abundance have frequently been underestimated in the past, and this problem persists. When two groups use the same data and different ways to estimate the electron density and 4He abundance, the results differ by more than the quoted systematic errors. While the methods used by Izotov & Thuan [1] seem to be an advance on those used before, the other method is reasonable, and hence the systematic error should encompass the range in results.

The abundance of 7Li is measured to high accuracy, but we do not know how much was produced prior to the formation of the stars, and how much was destroyed (depleted) in the stars. 6Li helps limit the amount of depletion of 7Li , but by an uncertain amount since it too has been depleted.

BBN is successful because it uses known physics and measured cross-sections for the nuclear reactions. It gives accurate predictions for the abundances of five light nuclei as a function of the one free parameter eta . The other initial conditions seem natural: the universe began homogeneous and hotter than T > 1011 K (30 Mev). The predicted abundances agree with most observations, and the required eta is consistent with other, less accurate, measurements of the baryon density.

New measurements of the baryon density, from the CMB, clusters of galaxies and the Lyalpha forest, will give eta . Although the accuracy might not exceed that obtained from D/H, this is an important advance because BBN then gives abundance predictions with no adjustable parameters.

New measurement in the coming years will give improved accuracy. Measurement of D/H in many more quasar spectra would improve the accuracy of D/H by a factor of a few, to a few percent, but even with improved methods of selecting the target quasars, this would need much more time on the largest telescopes. More reliable 4He abundances might be obtained from spectra which have higher spectral and spatial resolution, to help correct for stellar absorption, higher signal to noise to show weaker emission lines, and more galaxies with low metal abundances, to minimize the extrapolation to primordial abundances. Measurements of 6Li, Be and Boron in the same stars and observations of a variety of stars should give improved models for the depletion of 7Li in halo stars, and hence tighter constraints on the primordial abundance. However, in general, it is hard to think of any new methods which could give any primordial abundances with an order of magnitude higher accuracy than those used today. This is a major unexploited opportunity, because it means that we can not yet test BBN to the accuracy of the predictions.

Table of Contents