HeII emitters at cosmic noon and beyond. Characterising the HeII λ1640 emission with MUSE and JWST/NIRSpec

TL;DR

The paper investigates the origin of nebular He II 1640 Å emission in a sample of He II emitters at cosmic noon (z ~ 2–4) by combining MUSE rest-frame UV data with JWST/NIRSpec optical spectroscopy. Using UV–optical diagnostics, they derive very low gas-phase metallicities ($12 + ext{log(O/H)}$ between $7.3$ and $7.7$), sub-solar C/O, and roughly constant N/O, along with electron densities of $n_e \\sim 10^2$–$10^3$ cm^-3. Comparison with BPASS binary-star models and Gutkin UV photoionisation models shows that, for higher mass and metallicity, nearly metal-free single stars or very low-metallicity interacting binaries can reproduce the He II ionising budget, while lower-mass, very low-metallicity systems require top-heavy IMFs. Four extreme objects in the sample demand either nearly metal-free binaries or very low-metallicity binaries with top-heavy IMFs, indicating a persistent challenge for standard stellar populations to account for He II emission in extremely metal-poor regimes. The study demonstrates a clear stellar-mass–metallicity–multiplicity interplay in He II ionisation and motivates a larger, homogeneous sample to robustly constrain the drivers of He II emission and potential LyC leakage at high redshift.

Abstract

The study of high-redshift galaxies provides critical insights into the early stages of cosmic evolution, particularly during the so-called 'cosmic noon', when star formation activity reached its peak. Within this context, the origin of the nebular emission remains an open question. In this work, we conduct a systematic, multi-wavelength investigation of a sample of z ~ 2-4 emitters from the MUSE Hubble Ultra Deep Field surveys, utilising both MUSE and JWST/NIRSpec data and extending the sample presented by previous studies. We derive gas-phase metallicities and key physical properties, including electron densities, temperatures and the production rates of hydrogen- and He+-ionising photons. Our results suggest that a combination of factors-such as stellar mass, initial mass function, stellar metallicity, and stellar multiplicity-likely contributes to the origin of the observed nebular emission. Specifically, for our galaxies with higher gas-phase metallicity (12 + log(O/H) > 7.55), we find that models for binary population with Salpeter IMF (Mup=100 Msol) and stellar metallicity ~ 0.001 (i.e., similar to that of the gas) can reproduce the observed ionising conditions. However at lower metallicities, models for binary population with `top-heavy' initial mass function (Mup = 300 Msol) and Zstar much lower < Zstar) than that of the gas are required to fully account for the observed ionising photon production. These results reinforce that the ionisation keeps challenging current stellar populations, and the ionisation problem persists in the very low metallicity regime.

Figures (11)

• Figure 1: Example of a detected source from MHUDF, representing the galaxy with ID 106 in the MXDF. Upper panels: MUSE white-light image (left), ORIGIN segmentation map of the source (centre), and HST F606W filter image (right). Lower pannel: Observed reference spectrum of the source extracted with ORIGIN (blue) and its respective continuum + emission line fit (red).
• Figure 2: Stellar mass vs. star formation rate for the sample, derived from the SED reported in the AMUSED database MHXDFBacon23, colour-coded by redshift and excluding the two AGNs. The z = 3 SFMS, as derived by BoogaardSFMS18, is represented by a dashed blue line. The sample shows a slight offset relative to the z = 3 SFMS, but follows the expected linear trend.
• Figure 3: Reduced 2D and 1D NIRSpec spectrum of the galaxy with ID:50 at z = 3.325, obtained using the G235M/F170LP grating–filter combination. The right panels display a NIRCam RGB cutout of the galaxy, constructed using the F115W, F277W, and F444W filters (top), and a F444W cutout with the NIRSpec slit location overplotted in magenta (bottom). Both NIRCam cutouts were obtained from the grizli public database.
• Figure 4: He2-O3C3 diagnosis diagram. The red lines represent the polynomial from equation \ref{['eq:He2-O3C3']} at 12 + log$_{10}$(O/H) = 7.3, 7.5, and 7.7 respectively. The colours indicate stellar mass, while the open points correspond to galaxies without SED information, and consequently no measured $M_{\star}$. Only 18 points are shown as the rest present a S/N lower than 3 in at least one of the involved emission lines. All galaxies fall within a metal-poor regime; three of them approach the extremely metal-poor regime (12 + log(O/H) $\approx$ 7.3), showing a trend with $M_{\star}$.
• Figure 5: He$^{++}$ vs. H$^{+}$ ionising photon production, derived from Equations \ref{['eq:QHe']} and \ref{['eq:QH']}. The colours indicate the stellar mass of each galaxy. The black square represents measurements for the extremely metal-poor star-forming galaxy SBS 0335-052E taken from KehrigSBS18.