Helium Abundance in the Most Metal‐deficient Blue Compact Galaxies: I Zw 18 and SBS 0335−052

Open Access
Publisher Website
Google Scholar
Abstract
We present high-quality spectroscopic observations of the two most metal-deficient blue compact galaxies known, I Zw 18 and SBS 0335-052. We use the data to determine the heavy-element and helium abundances. The oxygen abundances in the northwest and the southeast components of I Zw 18 are found to be the same within the errors, 7.17 ± 0.03 and 7.18 ± 0.03, respectively, although marginally statistically significant spatial variations of oxygen abundance might be present. In contrast, we find a statistically significant gradient of oxygen abundance in SBS 0335-052. The largest oxygen abundance, 12 + log O/H = 7.338 ± 0.012, is found in the region 06 to the northeast of the brightest part of the galaxy, and it decreases toward the southwest to values of ~7.2, comparable to that in I Zw 18. The underlying stellar absorption strongly influences the observed intensities of He I emission lines in the brightest northwest component of I Zw 18, and hence this component should not be used for primordial He abundance determination. The effect of underlying stellar absorption, though present, is much smaller in the southeast component. Assuming all systematic uncertainties are negligible, the He mass fraction Y = 0.243 ± 0.007 derived in this component is in excellent agreement with recent measurements by Izotov & Thuan, suggesting the robustness of the technique applied in measurements of the helium abundance in low-metallicity blue compact galaxies. The high signal-to-noise ratio spectrum (≥100 in the continuum) of SBS 0335-052 allows us to measure the helium mass fraction with a precision better than 2%-5% in nine different regions along the slit. We show that, while underlying stellar absorption in SBS 0335-052 is important only for the He I 4471 Å emission line, other mechanisms such as collisional and fluorescent enhancements are influencing the intensities of all He I emission lines and should be properly taken into account. When the electron number density derived from [S II] emission lines is used in SBS 0335-052, the correction of He I emission lines for collisional enhancement leads to systematically different He mass fractions for different He I emission lines. This unphysical result implies that the use of the electron number density derived from [S II] emission lines, being characteristic of the S+ zone but not of the He+ zone, will lead to an incorrect inferred value of Y. In the case of SBS 0335-052 it leads to a significant underestimate of the He mass fraction. In contrast, the self-consistent method using the five strongest He I emission lines in the optical spectrum for correction for collisional and fluorescent enhancements shows excellent agreement of the He mass fraction derived from the He I 5876 Å and He I 6678 Å emission lines in all nine regions of SBS 0335-052 used for the He abundance determination. Assuming all systematic uncertainties are negligible, the weighted mean He mass fraction in SBS 0335-052 is Y = 0.2437 ± 0.0014 when the three He I 4471, 5876 and 6678 Å emission lines are used, and it is 0.2463 ± 0.0015 when the He I 4471 Å emission line is excluded. These values are in very good agreement with recent measurements of the He mass fraction in SBS 0335-052 by Izotov and coworkers. The weighted mean helium mass fraction in the two most metal-deficient blue compact galaxies, I Zw 18 and SBS 0335-052, Y = 0.2462 ± 0.0015, after correction for the stellar He production results in a primordial He mass fraction Yp = 0.2452 ± 0.0015. The derived Yp leads to a baryon-to-photon ratio of 4.7 × 10-10 and to a baryon mass fraction in the universe Ωbh = 0.068, consistent with the values derived from the primordial D and 7Li abundances, and supporting the standard big bang nucleosynthesis theory. For the most consistent set of primordial D, 4He, and 7Li abundances we derive an equivalent number of light neutrino species Nν = 3.0 ± 0.3 (2 σ).
All Related Versions

This publication has 60 references indexed in Scilit: