The primordial abundance of 4He: evidence for non-standard big bang nucleosynthesis
Y. I. Izotov, T. X. Thuan
TL;DR
This work delivers a refined determination of the primordial helium mass fraction $Y_p$ by analyzing 93 spectra from 86 low-metallicity H II regions and rigorously accounting for major systematic effects via a Monte Carlo approach. The authors model the relation between helium mass fraction and metallicity with $Y = Y_p + (dY/d({\rm O/H}))({\rm O/H})$, and derive $y^+$ and other quantities through a simultaneous, statistically robust treatment of line emissivities, reddening, temperature structure, and ionization corrections. They obtain $Y_p \approx 0.2565$ (stat) and a substantial systematic error, and find a slope $dY/dZ \approx 1.6$, indicating a higher primordial value than standard BBN predictions. A joint cosmological-BBN analysis with deuterium and lithium data implies an effective number of light neutrino species $N_\nu \approx 3.7$–$3.8$, suggesting deviations from standard BBN and potential new physics in the early universe.
Abstract
We present a new determination of the primordial helium mass fraction Yp, based on 93 spectra of 86 low-metallicity extragalactic HII regions, and taking into account the latest developments concerning systematic effects. These include collisional and fluorescent enhancements of HeI recombination lines, underlying HeI stellar absorption lines, collisional and fluorescent excitation of hydrogen lines and temperature and ionization structure of the HII region. Using Monte Carlo methods to solve simultaneously for the above systematic effects, we find the best value to be Yp=0.2565+/-0.0010(stat.)+/-0.0050(syst.). This value is higher at the 2sigma level than the value given by Standard Big Bang Nucleosynthesis (SBBN), implying deviations from it. The effective number of light neutrino species Nnu is equal to 3.68^+0.80_-0.70 (2sigma) and 3.80^+0.80_-0.70 (2sigma) for a neutron lifetime tau(n) equal to 885.4+/-0.9 s and 878.5+/-0.8 s, respectively, i.e. it is larger than the experimental value of 2.993+/-0.011.