An Ultra-Faint, Chemically Primitive Galaxy Forming in the Reionization Era

Abstract

The formation of the first stars and galaxies marked the onset of chemical enrichment, yet direct observations of such primordial systems remain elusive. Here we present James Webb Space Telescope spectroscopic observations of LAP1-B, an ultra-faint galaxy at redshift z_{spec}=6.625 +/-0.001, corresponding to a cosmic age of 800 million years after the Big Bang, strongly magnified by gravitational lensing. LAP1-B exhibits a gas-phase oxygen abundance of (4.2 +/- 1.8) x 10^{-3} times the solar value, making it the most chemically primitive star-forming galaxy discovered to date. The galaxy displays an exceptionally hard ionizing radiation field, which is inconsistent with chemically enriched stellar populations or accreting black holes but matches theoretical predictions for an exceptionally metal-deficient stellar population. It also shows an elevated carbon-to-oxygen abundance ratio for its metallicity in the interstellar medium, consistent with nucleosynthetic yields from a stellar population formed in the absence of initial metals. The lack of detectable stellar continuum constrains the stellar mass to <3,300 Msun, while the dynamical mass, derived from emission-line kinematics, exceeds the combined stellar and gas mass and indicates a dominant dark matter halo. Our findings establish LAP1-B as a "fossil in the making", a direct high-redshift progenitor of the ancient ultra-faint dwarf galaxies observed in the local Universe, offering a rare window into the earliest stages of galaxy formation.

Figures (9)

• Figure 1: NIRCam image and NIRSpec spectra of LAP1-B. (a Left:) False-color NIRCam image of the MACS J0416 cluster, combining F115W and F150W (blue), F200W and F277W (green), and F356W and F444W (red). The inset shows a 2.$\prime\prime$5$\times$2.$\prime\prime$5 zoom-in of the full-composite image (stacked across F115W to F444W) centered on the location of LAP1-B (indicated with white lines), where no counterpart is detected. The two footprints of the NIRSpec 3-shutter slitlets used for spectroscopy are overlaid in red and orange. (b Right:) JWST/NIRSpec spectra of LAP1-B centered on the emission lines discussed in this work. In each panel, the top shows the 2D spectrum, and the bottom shows the extracted 1D spectrum (black) with the $1\sigma$ uncertainty shaded. The expected wavelengths of emission lines at the systemic redshift $z = 6.625$ are marked by vertical dashed lines. Emission signals detected within $\pm 1.5\times$ the instrumental FWHM are highlighted in red.
• Figure 1: Oxygen abundance diagnostics based on [O iii]$\lambda 5007$/H$\beta$. The relationship between the [O iii]$\lambda 5007$/H$\beta$ line ratio and gas-phase oxygen abundance is shown. Curves with symbols represent empirically calibrated relations derived from galaxies at $z=2-9$sanders2024, chakraborty2025, extremely metal-poor galaxies at $z\simeq 0$ with EW(H$\beta$) $\geq 200$ Å nakajima2022_empressV, and cosmological simulations hirschmann2023, as indicated in the legend. Each curve is plotted over the range explored in the respective studies. Overlaid are theoretical predictions from photoionization models NM2022, using two types of stellar ionizing sources: chemically enriched BPASS stellar populations (continuous star formation history over 10 Myr with a Kroupa IMF up to 300 $M_{\odot}$) and zero-metallicity Population III stars (Salpeter IMF, 1-100 $M_{\odot}$schaerer2002). Different curves correspond to varying ionization parameters: from $\log U = -0.5$ to $-2.0$ for the Population III models, and extending down to $\log U = -3.5$ for the BPASS models. The Population III model with the highest ionization parameter is adopted to derive the oxygen abundance of LAP1-B (see text for justification), whose observed [O iii]/H$\beta$ ratio and derived oxygen abundance are indicated in red.
• Figure 2: Chemical enrichment as a function of stellar mass across cosmic time. The oxygen abundance and stellar mass of LAP1-B are shown with two symbols. The filled red circle represents our fiducial measurement, while the open red circle indicates the conservative upper limit on mass, as discussed in the text. This measurement is compared with those of $300$ other galaxies at $z=4$--$12.5$ identified with JWST (brown symbols), whose metallicities were derived in a comparable manner. For local comparison, we show $z=0$ UFDs and more massive counterparts of UFDs, referred to as classical dwarf spheroidal galaxies (dSphs) as grey crosses (individual systems) and a grey dotted line (average relation). Their metallicities are based on stellar [Fe/H] measurements derived from the metallicity distributions of individual stars. These values are converted to oxygen abundances assuming [O/Fe] $=+0.5$. We note this is an approximate comparison; the present-day stellar masses of these dwarf galaxies can be lower limits on their progenitor masses due to stellar evolutionary mass loss, and the precise behavior of the [O/Fe] ratio in such metal-poor environments is a subject of ongoing research. References for the data compiled here are provided in the Methods section.
• Figure 2: Diagnostic diagrams probing the shape of ionizing spectrum. LAP1-B (red) is compared with photoionization model predictions NM2022 and some from extremely metal-poor models from this work using two diagnostic plots: EW(He ii$\lambda 1640$) versus He ii/H$\beta$ line ratio (left), and the ionizing photon production efficiency ($\xi_{\rm ion}$) versus EW(H$\alpha$) (right). $\xi_{\rm ion}$ is derived as the H$\alpha$ luminosity divided by the UV continuum flux density, with the resulting values plotted on the right-hand y-axis. The measurement for LAP1-B conservatively assumes a zero escape fraction of ionizing photons; this provides a lower limit, as a non-zero escape fraction would imply an even higher intrinsic $\xi_{\rm ion}$. Model predictions shown in both panels correspond to different ionizing sources: Population III stars, chemically enriched Population II stars, and accreting black holes (see ref NM2022 for details). Symbol shapes distinguish between these ionizing sources, while colors indicate different gas-phase metallicities, as in the legend. Note that the newly developed $Z=10^{-7}$ Population II models with a $50$--$500$$M_{\odot}$ IMF overlap the Population III region in the right-hand $\xi_{\rm ion}$ panel, offering an alternative explanation for the observed value.
• Figure 3: Comparison of observed emission-line ratios with photoionization models. The observed line ratios for LAP1-B (red circle) are compared to photoionization model predictions NM2022 in the C iv/[O iii] versus [O iii]/H$\beta$ diagnostic diagram. The top row contrasts fiducial models: a zero-metallicity Population III population (left) versus a chemically enriched Population II population with standard assumptions (matched stellar and gas-phase metallicities, Kroupa IMF; right). The bottom row explores variations of the Population II models to test the impact of IMF and dust depletion. The first two bottom panels assume an extremely low stellar metallicity ($Z_{\text{star}}=10^{-7}$) and compare a standard IMF ($1$--$100$$M_{\odot}$; left) to one composed of very massive stars ($50$--$500$$M_{\odot}$; center). The bottom-right panel tests the role of dust depletion by assuming zero depletion for two scenarios: the fiducial Population II model ($Z_{\text{star}} = Z_{\text{gas}}$; long-dashed with shade) and the extreme top-heavy model ($Z_{\text{star}} = 10^{-7}, 50$--$500$$M_{\odot}$; solid lines). Each panel illustrates model grids for three C/O abundance ratios: $2\times$ solar (green), solar (blue), and an empirically motivated relation (grey) corresponding to $\sim 0.2\times$ solar at these metallicities. Grids connect points of constant oxygen abundance and ionization parameter, with values indicated in the legend.