A Pristine Star-Forming Complex at z=4.19

TL;DR

This paper reports the discovery of LAP2, a faint, oxygen-deficient star-forming complex at $z=4.19$ strongly lensed by Abell 2744. JWST/NIRSpec prism spectroscopy detects Ly$\alpha$ and H$\alpha$ but not [OIII], yielding a stringent metallicity limit of $Z<6\times10^{-3}\,Z_{\odot}$ and $12+\log(\mathrm{O/H})<6.5$, while lensing reveals a sub-10 pc size and an intrinsic UV magnitude of $M_{\rm UV}\approx -12.2$. The data imply an instantaneous-burst, very young stellar population with $M_{*} \sim$ a few $\times 10^{4}\,M_{\odot}$ and a high ionizing photon production efficiency $\log_{10}(\xi_{\rm ion}) \gtrsim 25.13$, placing LAP2-b among the most metal-poor star-forming systems known at $z>3$. The copious Ly$\alpha$ emission observed by VLT/MUSE as an arclet near the critical line, together with planned deeper NIRSpec/IFU follow-up, will test pristine star-formation scenarios and extend metallicity studies into the faint-luminosity regime.

Abstract

We report the discovery of a faint (M_1700 ~ -12.2), oxygen-deficient strongly-lensed ionizing source -- dubbed LAP2 (Lensed And Pristine 2) -- at a spectroscopic redshift of z=4.19. LAP2 appears to be isolated and lies very close to the caustic produced by the lensing galaxy cluster Abell 2744. It was observed with the James Webb Space Telescope (JWST) NIRSpec MSA in prism mode as part of the UNCOVER program. The NIRSpec spectrum reveals prominent Lya emission (7.1 sigma), clear Ha emission (6.2 sigma), tentative Hb emission (2.8 sigma), and no detectable [OIII]4959,5007 (~ 7 times fainter than Ha). The inferred [OIII] 2 sigma upper limit corresponds to an R3 index <0.85 (assuming the Ha/Hb ~ 2.86 case~B recombination ratio), which, under high-ionization conditions, implies a metallicity of Z < 6 X 10^(-3) Z_sun. The combination of faint ultraviolet luminosity, large rest-frame Ha equivalent width (~ 650 A), and extremely compact size (< 10 pc) suggests that LAP2 is being caught in an early, pristine formation phase consistent with an instantaneous-burst scenario, with an estimated stellar mass of at most a few 10^4 Msun. Deep VLT/MUSE observations further reveal copious Lya emission forming an arclet that straddles the critical line. LAP2 thus joins the rare class of extremely metal-poor star-forming complexes that JWST has started to unveil at redshifts 3 - 7, and provides a rare glimpse into a still very poorly explored faint-luminosity regime.

Figures (3)

• Figure 1: Schematic view of α Ly$\alpha$ emitter at $z=4.189$ amplified by the galaxy cluster A2744 (background color image, F090W, F150W, F200W). The yellow contours mark $\mu=50$ at $z=4.19$Bergamini2023. The NIRSpec MSA pointing (green box) is shown in the bottom left inset, which zooms in on the α Ly$\alpha$ arclet and its multiple images, dubbed LAP2-a,b (the cyan lines outline the 3 and 8 $\sigma$ α Ly$\alpha$ contour from VLT/MUSE). The upper right inset shows the VLT/MUSE α Ly$\alpha$ line emission of LAP2-b. The bottom right inset shows the NIRSpec prism spectrum and error of LAP2-b in F$_\nu$ units with the fluxes inferred from the spectrum convolved with NIRCam filters (open squares) and the corresponding measured NIRCam magnitudes indicated.
• Figure 2: Overview of JWST/NIRSpec MSA pointing on LAP2. The 2D spectrum is shown on the top with the more relevant lines indicated, along with the area used for the extraction of the spectrum (dashed horizontal lines). The yellow shaded area marks contamination signal arising from a nearby source, and the blue shaded area on the left marks the wavelength coverage of VLT/MUSE. The bottom left and middle insets show the line and continuum fitting (dashed colored curves). The light gray shaded area indicates the flux uncertainties, while the light red area in the bottom middle panel indicates the expected wavelength of the [Oiii] doublet. Both the 1D and 2D spectra reveal the absence of oxygen emission. The bottom right inset shows a zoomed-in view ($1.9"\times1.6 "$) of the F090W band of LAP2-b targeted with MSA (outlined with the black boxes). The blue contours mark the VLT/MUSE α Ly$\alpha$ at 3 and 8 sigma, along with the locus of $\mu=50$ (red line).
• Figure 3: Oxygen abundance of LAP2-b, as function of its R3 index (left panel, 2$\sigma$ limit is reported) and its stellar mass (right panel). Additional measurements of low-metallicity sources at $z>3$ from the literature are reported Vanzella2023_lap1Morishita2024Chemerynska2024Mowla2024Tripodi2025Vanzella2024_t2cHsiao2025Curti2025Cullen2025Willott2025Maiolino2025Nakajima2025Cai2025Morishita2025. Samples plotted with empty markers have their metallicity estimated from the direct $T_e$ method, as opposed to the ones derived from strong line ratios (filled markers). The two most common R3-to-Z calibrations used in high-$z$ studies as well as the conversion based on the BPASS models used in Morishita2025 (dotted black line), are shown in the left panel (see legend). Two recent mass-metallicity relations derived from high-z samples (blue line: Curti2024; black line: Morishita2024) are shown in the right panel; the mass ranges used to derive the relations are shown as the solid portion of the line.