Precision measures of the primordial abundance of deuterium

TL;DR

This work delivers a new precise measurement of the primordial deuterium abundance from a very metal-poor DLA and expands the dataset by reanalyzing four additional DLAs with a uniform, blind methodology. The combined Precision Sample yields (D/H)$_p = (2.53 ± 0.04)×10^{-5}$, translating to 100$\Omega_{\rm b,0}h^2 = 2.202 ± 0.046$, in excellent agreement with Planck CMB results. By integrating (D/H)$_p$ with Planck data, the paper tightens constraints on the effective number of neutrino species, $N_{\rm eff} = 3.28 ± 0.28$, and places limits on the neutrino degeneracy parameter, $|\xi| ≤ 0.062$ (2σ), thereby disfavoring additional sterile neutrinos under standard assumptions. The study also identifies the dominant limitation as uncertainties in key nuclear reaction cross sections, notably d(p,γ)³He, and calls for improved measurements and more high-quality DLAs to further test cosmology and new physics.

Abstract

We report the discovery of deuterium absorption in the very metal-poor ([Fe/H] = -2.88) damped Lyman-alpha system at z_abs = 3.06726 toward the QSO SDSS J1358+6522. On the basis of 13 resolved D I absorption lines and the damping wings of the H I Lyman alpha transition, we have obtained a new, precise measure of the primordial abundance of deuterium. Furthermore, to bolster the present statistics of precision D/H measures, we have reanalyzed all of the known deuterium absorption-line systems that satisfy a set of strict criteria. We have adopted a blind analysis strategy (to remove human bias), and developed a software package that is specifically designed for precision D/H abundance measurements. For this reanalyzed sample of systems, we obtain a weighted mean of (D/H)_p = (2.53 +/- 0.04) x 10^-5, corresponding to a Universal baryon density100 Omega_b h^2 = 2.202 +/- 0.046 for the standard model of Big Bang Nucleosynthesis. By combining our measure of (D/H)_p with observations of the cosmic microwave background, we derive the effective number of light fermion species, N_eff = 3.28 +/- 0.28. We therefore rule out the existence of an additional (sterile) neutrino (i.e. N_eff = 4.046) at 99.3 percent confidence (2.7 sigma), provided that N_eff and the baryon-to-photon ratio (eta_10) did not change between BBN and recombination. We also place a strong bound on the neutrino degeneracy parameter, xi_D = +0.05 +/- 0.13 based only on the CMB+(D/H)_p observations. Combining xi_D with the current best literature measure of Y_p, we find |xi| <= +0.062. In future, improved measurements of several key reaction rates, in particular d(p,gamma)3He, and further measures of (D/H)_p with a precision comparable to those considered here, should allow even more stringent limits to be placed on new physics beyond the standard model.

Figures (10)

• Figure 1: Flux-calibrated spectrum of J1358$+$6522 (black histogram). The zero-level and error spectrum are shown as the green dashed and solid blue lines respectively. The Hi Lyman series lines of the DLA are indicated by red tick marks. Since the SDSS data do not extend below 3800 Å, we extrapolated the response curve for fluxing the data a further 100 Å to the blue.
• Figure 2: The top panel displays a portion of the flux-calibrated HIRES spectrum near the damped Ly$\alpha$ line at $z_{\rm abs}=3.06726$ toward J1358$+$6522 (black histogram) together with the error spectrum (continuous blue line). The dashed green line marks the best-fitting zero level of the data and the dashed blue line shows the best-fitting continuum level. The solid red line shows the overall best-fitting model to the DLA. The bottom panel shows a zoom-in of the data and model; the weak absorption feature that we have modelled on the blue wing of Ly$\alpha$ is Si iii$\lambda 1206.5$ at the redshift of the DLA ($\sim4907$ Å in the observed frame).
• Figure 3: A montage of the full Lyman series absorption in the DLA at $z_{\rm abs}=3.067259$ toward J1358$+$6522. The black histogram shows the data, fully adjusted to the best-fitting continuum and zero levels, while the red continuous line is the model fit. The minimum $\chi^{2}$/dof for this fit is $6282.3/6401$. Tick marks above the spectrum indicate the location of the velocity components (red ticks for Hi, green ticks for Di).
• Figure 4: A selection of metal absorption lines for the DLA at $z_{\rm abs}=3.067259$ toward the QSO J1358$+$6522 (black histogram, cf. Coo12). The best-fitting model is shown as the solid red line, while fitted blends are shown as the thick orange lines. In the top panels, the red tick marks above the spectrum indicate the location of the two absorption components, while the bottom panels only display the tick mark for the main absorption component (i.e. component 1a+1b). The long-dashed blue and dashed green lines show the continuum and zero level respectively. The absorption feature near $+25$ km s$^{-1}$ in the Si ii$\lambda1260$ panel is Fe ii$\lambda1260.5$.
• Figure 5: Values of D/H for the Precision Sample of DLA measurements analyzed in this paper. The orange point represents the new case reported here (J1358+6522). The left and right panels show respectively the D/H measures as a function of the DLA oxygen abundance and Hi column density. The dark and light green bands are the 1$\sigma$ and 2$\sigma$ determinations of $\Omega_{\rm b,0}$$\,h^2$ from the analysis of the CMB temperature fluctuations recorded by the Planck satellite Efs13 assuming the standard model of physics. The conversion from D/H to $\Omega_{\rm b,0}$$\,h^2$ is given by eqs. \ref{['eqn:dhptoetad']} and \ref{['eqn:etad']}.