1.3. Non-standard BBNS models

In recent years there has been active discussion of alternative, non-standard BBNS models that postulate baryon density fluctuations arising from the quark-hadron phase transition (if this is indeed first-order) and related variations in the n / p ratio due to differential diffusion of protons and neutrons (Applegate & Hogan 1985; Alcock, Fuller & Mathews 1987; Applegate, Hogan & Scherrer 1988; Kawano, Fowler & Malaney 1990). In particular, it has been suggested that such models could fit light-element abundances with _b = 1. The analysis involves a number of free parameters (density contrast, filling factors and length scales) and the proper treatment of all diffusion effects is difficult and controversial, but at the present time models with _b = 1 do not seem to be viable because they predict too much helium and lithium 7 for any combination of the free parameters (Terasawa & Sato 1989, 1990; Reeves 1988, 1990; Kurki-Suonio et al. 1990). However, Reeves (1990) and Kurki-Suonio et al. find that mildly inhomogeneous models are quite plausible and could fit the data for somewhat wider limits on than are given by SBBN; a rough adaptation of these wider limits from the latter reference is shown by the shorter double vertical lines in fig. 1. Upper limits on N, and _1/2 are affected very little in these models. A remote possibility exists that there might be significant primordial abundances of elements above ⁷Li from some kind of inhomogeneous BBNS, but existing data certainly do not suggest anything of the sort (Pagel 1991).

A completely different type of non-standard BBNS theory involves hypothetical massive, unstable particles (e.g. photinos, massive neutrinos, antimatter etc.) which could have various effects depending on their mass, interaction strength and lifetime. For example, they could modify the equation of state, and the success of SBBN and accelerator experiments now rule out large regions of parameter space. They might also decay before, during or after BBNS, modifying the final products. Dimopoulos et al. (1988) suggested that massive (> 2 Gev) particles decaying after 10⁵ s (early enough not to disturb the microwave background) produce electromagnetic and hadron showers which wipe the slate clean after BBNS and remove SBBN restrictions on _b and N. This particular model makes detailed predictions that disagree with astrophysical observations, in particular too high a ratio of ⁶Li to ⁷Li (Audouze & Silk 1989), but these models are generically unappealing on the more fundamental grounds that they perversely throw away the impressive predictions of SBBN theory.