Table of Contents
Fetching ...

Nitrogen abundances in star-forming galaxies 2.2 Gyr after the Big Bang are not elevated

D. Schaerer, Y. I. Izotov, R. Marques-Chaves, C. C. Steidel, N. Reddy, A. E. Shapley, S. Mascia, J. Chisholm, S. R. Flury, N. Guseva, T. Heckman, A. Henry, A. K. Inoue, I. Jung, H. Kusakabe, K. Mawatari, P. Oesch, G. Oestlin, L. Pentericci, N. Roy, A. Saldana-Lopez, R. Sato, E. Vanzella, A. Verhamme, B. Wang

TL;DR

This study uses deep JWST/NIRSpec rest-optical spectra of LyC22 galaxies at $z\sim3$ to derive nebular abundances via the direct method for 25 objects. The measured $\\log({\rm N/O}) = -1.29^{+0.25}_{-0.22}$ and $12+\log({\rm O/H})$ ranging from 7.56 to 8.44 show nitrogen abundances that are not elevated compared to local H II regions, indicating no strong redshift evolution of N/O for typical galaxies over a substantial fraction of cosmic time. The Ne/O ratio is $\\log({\rm Ne/O}) = -0.73^{+0.11}_{-0.08}$ and the galaxies display BPT offsets similar to $z\sim2.3$ systems, with the offset not caused by enhanced nitrogen. These results constrain the chemical evolution of nitrogen and oxygen at high redshift and establish a baseline for comparing to even earlier epochs ($z>4$) as larger samples become available.

Abstract

Using deep medium-resolution JWST rest-optical spectra of a sample of typical star-forming galaxies (Lyman break galaxies and Lyman-$α$ emitters) from the LyC22 survey at $z \sim 3$, we determined the nebular abundances of N, O, and Ne relative to H for a subsample of 25 objects with the direct method, based on auroral [OIII]4363 line detections. Our measurements increases the number of accurate N/O determinations at $z \sim 2-4$ using a homogeneous approach. We found a mean value of $\log({\rm N/O})=-1.29^{+0.25}_{-0.21} $ over a metallicity range 12+log(O/H)=7.5 to 8.44. The observed N/O ratio and scatter are indistinguishable from that observed in low-z galaxies and HII regions over the same metallicity range, showing thus no redshift evolution of N/O for typical galaxies over a significant fraction of cosmic time. We also show that typical $z \sim 3$ galaxies show a similar offset in the BPT diagram as galaxies from the low-z Lyman Continuum Survey (LzLCS), when compared to the average of SDSS galaxies, and show that this offset is not due to enhanced nitrogen abundances. Our results establish a basis for future studies of the evolution of N and O at higher redshifts.

Nitrogen abundances in star-forming galaxies 2.2 Gyr after the Big Bang are not elevated

TL;DR

This study uses deep JWST/NIRSpec rest-optical spectra of LyC22 galaxies at to derive nebular abundances via the direct method for 25 objects. The measured and ranging from 7.56 to 8.44 show nitrogen abundances that are not elevated compared to local H II regions, indicating no strong redshift evolution of N/O for typical galaxies over a substantial fraction of cosmic time. The Ne/O ratio is and the galaxies display BPT offsets similar to systems, with the offset not caused by enhanced nitrogen. These results constrain the chemical evolution of nitrogen and oxygen at high redshift and establish a baseline for comparing to even earlier epochs () as larger samples become available.

Abstract

Using deep medium-resolution JWST rest-optical spectra of a sample of typical star-forming galaxies (Lyman break galaxies and Lyman- emitters) from the LyC22 survey at , we determined the nebular abundances of N, O, and Ne relative to H for a subsample of 25 objects with the direct method, based on auroral [OIII]4363 line detections. Our measurements increases the number of accurate N/O determinations at using a homogeneous approach. We found a mean value of over a metallicity range 12+log(O/H)=7.5 to 8.44. The observed N/O ratio and scatter are indistinguishable from that observed in low-z galaxies and HII regions over the same metallicity range, showing thus no redshift evolution of N/O for typical galaxies over a significant fraction of cosmic time. We also show that typical galaxies show a similar offset in the BPT diagram as galaxies from the low-z Lyman Continuum Survey (LzLCS), when compared to the average of SDSS galaxies, and show that this offset is not due to enhanced nitrogen abundances. Our results establish a basis for future studies of the evolution of N and O at higher redshifts.
Paper Structure (10 sections, 5 figures)

This paper contains 10 sections, 5 figures.

Figures (5)

  • Figure 1: JWST spectra from LyC22 showing object 1005 (top panel, LACES $z=3.676$ LAE in SSA22) and 300009 (bottom, $z=2.767$ LBG in the Westphal field). The blue line shows the adopted fit to the continuum. The main emission and absorption lines are indicated.
  • Figure 2: Comparison of the N/O abundances derived following the direct method and the assumptions described in this work (x-axis) with those adopting the strong line calibrations of Cataldi2025Tracing-Nitroge based on JWST samples (in black; applicable for $z>1$) and an SDSS sample (red, applicable at low-$z$), and the calibration from Strom2017Nebular-Emissio (blue). All strong line calibrations use the extinction-corrected line ratio of [N ii] $\lambda$6584/[O ii] $\lambda\lambda$3727,3729.
  • Figure 3: Left: Classical emission line diagnostic diagram showing the galaxies from the LyC22 sample (blue and red symbols). LyC22 objects with significant [O iii] $\lambda 4363$ detections are surrounded by a black square. LyC22 galaxies with $\log({\rm N/O})>-1.2$ are shown in red. We only show objects where the involved emission lines are detected at $\ge 3 \sigma$. Typical uncertainties are comparable to the size of the symbols, hence not plotted. The maximum starburst line from Kewley2001Optical-Classif (dotted) and the empirical AGN/star-formation threshold from Kauffmann03 established for $z \sim 0$ (dash-dotted) are also shown. Average relations for $z \sim 2.3$ galaxies from Steidel2014Strong-Nebular-Shapley2015The-MOSDEF-Surv are shown by orange and magenta lines, average relation of SDSS galaxies from Kewley2013The-Cosmic-BPT- by the red line. Observations from the LzLCS at $z\sim 0.3$ are shown by small grey squares. Right: Zoom on part of the BPT diagram showing all LyC22 sources with color-coded N/O abundances (in $\log($N/O$)$), computed from [N ii]/[O ii].
  • Figure 4: Derived abundance ratio N/O from rest-optical lines of the LyC22 galaxies (blue circles) and low-$z$ samples as a function of O/H. The low-$z$ star-forming galaxies and H ii regions from the compilation of Izotov2023Abundances-of-C and the LzLCS are shown by small black and grey symbols, respectively. Dash-dotted and dotted lines show the average trend observed in low-$z$ star-forming galaxies, as parametrized by Vila-Costas1993The-nitrogen-to and Nicholls2017Abundance-scali respectively. The green dotted line shows the solar value.
  • Figure 5: Excess $\Delta(\log({\rm N/O}))$ of N/O with respect to the local N/O--O/H relation from Vila-Costas1993The-nitrogen-to as a function of redshift for galaxies with abundances from the direct method and N/O from rest-optical lines. Low-$z$ star-forming galaxies and H ii regions, LzLCS and LyC22 samples are shown with the same symbols as in Fig. \ref{['fig_abund']}. Other N/O measurements from the rest-optical lines, obtained primarily from JWST spectra, are shown at $z \sim 2.2$Sanders2023A-Preview-of-JW, for the EXCELS sample Arellano-Cordova2024The-JWST-EXCELSScholte2025The-JWST-EXCELS, from Stiavelli2025What-Can-We-Lea, the Sunburst arc at $z=2.4$Welch2025The-Sunburst-Ar, and for the $z=6.1$ lensed galaxy (orange) from Topping2024Metal-poor-star, with measurements from rest-optical lines by Berg2025A-Fleeting-GLIM. The two orange symbols indicate objects known to show also UV emission lines of Nitrogen.