No Metallicity Preference in Fast Radio Burst Host Galaxies

TL;DR

This study provides the largest uniform analysis of gas-phase metallicity in FRB host galaxies to date, assessing whether FRB progenitors prefer particular metallicity environments. Using a volume-limited, star-forming galaxy comparison and consistent O3N2-based metallicities, FRB hosts broadly trace the star-formation-weighted mass–metallicity relation across 0.01 < z < 0.48, with no sharp metallicity cutoff observed. A marginal offset below the fundamental metallicity relation is found, but is sensitive to calibration choices and could reflect modestly suppressed star formation in some hosts rather than intrinsic metal deficiency. The results imply FRBs can form in a wide range of metallicities, including low-metallicity, high-redshift systems, reinforcing their potential as probes of cosmic baryons and metal evolution, while highlighting possible post-starburst progenitor channels in some hosts.

Abstract

Fast radio bursts (FRBs) are millisecond-duration extragalactic radio transients of unknown origin, and studying their host galaxies could offer clues to constrain progenitor models. Among host properties, gas-phase metallicity is a key factor influencing stellar evolution and transient productions. We analyze the largest uniformly selected sample of FRB host galaxies, measuring oxygen abundances (12+log(O/H)=8.04-8.85) for 31 hosts at redshifts z=0.04-0.98, using consistent emission-line diagnostics. Using a volume-limited subsample, we compare the distributions of stellar mass, star formation rate (SFR), and metallicity to a control sample of star-forming galaxies selected by the same criteria. We find that FRB host galaxies span a wide metallicity range and are broadly consistent with the SFR-weighted mass-metallicity relation of star-forming galaxies. We find no clear lower metallicity bound, suggesting that FRB progenitors can form in any metallicity environment through channels largely insensitive to metal abundance. Encouragingly, this implies FRBs can arise even in low-metallicity, high-redshift galaxies, supporting their potential as probes of matter distribution across cosmic time. Additionally, we find marginal ($\sim$2$σ$) evidence for a -0.09\pm0.04 dex metallicity offset from the fundamental metallicity relation. Despite model uncertainties, if real, this offset likely reflects suppressed SFRs at fixed mass and metallicity rather than metal deficiency. Similar offsets are observed in local post-merger galaxies and may reflect a post-starburst phase following galaxy interactions. Such systems may host FRB progenitors formed during the starburst that produce FRBs after a 100-500 Myr delay, broadly consistent with observed delay-time distributions, although further data are needed to confirm this interpretation.

Figures (7)

• Figure 1: Host galaxy selection following Steps 2 and 3 of the "volume-limited" sample described in Table \ref{['tab:selection']}. Step 2 ( left): Volume-limited selection in the redshift–absolute r-band magnitude plane. The solid curve shows the apparent magnitude limit of m$_r=23.5$, used as a reference to indicate the approximate depth of the host galaxy samples based on the completeness limit of the DSA sample in sharma24. The gray shaded region indicates our selection window, optimized to include the maximum number of sources (light blue filled circles). Six sources (open circles) fall outside this window and are excluded. Step 3 ( right): BPT-diagram-based selection of star-forming galaxies. From the Step 1 sample, we select only those classified as SF galaxies based on emission line ratios to construct the final sample (dark blue filled circles). Black arrows indicate data points with upper or lower limits. The sample was compiled from eftekhari23gordon23sharma24.
• Figure 2: Comparison of FRB host galaxy properties with the general galaxy population. (a) Stellar mass vs. SFR ($N=29$; see Step 3 of "volume-limited" sample in Table \ref{['tab:selection']}), (b) Stellar mass vs. gas-phase metallicity ($N=27$, plus one upper limit; see Step 4 of "volume-limited" sample in Table \ref{['tab:selection']}), (c) Fundamental metallicity relation ($N=27$, plus one upper limit; same as panel b). FRB host galaxies are shown as open blue circles. Light green points represent the control sample. Green dashed, purple solid, and orange dotted contours show the original, SFR-weighted, and mass-weighted control sample distributions, respectively. Contours correspond to the 20%, 50%, and 80% cumulative density levels. The 2D-KS $p$-values are indicated in each panel.
• Figure 3: Comparison of cumulative distributions of FRB host galaxy properties (“volume-limited” sample defined in Table \ref{['tab:selection']} and used in Figure \ref{['fig:dist_2d']}): from left to right, gas-phase metallicity ($N=27$), stellar mass ($N=29$), and SFR ($N=29$). Solid blue histograms show the CDFs of the FRB hosts, and the shaded blue areas indicate the 95% confidence intervals estimated by bootstrapping. Colored curves show the different models discussed in the text: the unweighted model (dashed green), the SFR-weighted model (solid purple), and the mass-weighted model (dotted orange). The 1D-KS $p$-values are indicated in each panel.
• Figure 4: Left: FRB host galaxies ( $N=31$, plus 2 upper/lower limits) selected based on the BPT diagram (see "no-volume-limit" sample in Table \ref{['tab:selection']} and \ref{['tab:frb_host_data']} in Appendix \ref{['s:fullsample']} for the details), compared with the fundamental metallicity relation (FMR) from sanders21, which has been robustly confirmed for galaxies at $z=0$–3. The FMR is shown as a dashed black line with a dispersion of $0.05$ dex along the metallicity axis. Right: Histogram of metallicity offsets from the FMR for the FRB host sample, using only the $N=31$ hosts with metallicity measurements. The mean ($\mu$), standard deviation ($\sigma$) and standard error ($\sigma/\sqrt{N}$) are indicated in the panel.
• Figure 6: Median shift in the inferred values of $\log M_\star$ and $\log{\rm SFR}$ between the Prospector non-parametric SFH (npSFH) and parametric SFH (pSFH) models for galaxies at $0.7 < z < 1.3$ (from Fig. 9 of leja22). Only cells with $N > 20$ galaxies are shown. The $x$- and $y$-axes are defined in the npSFH scale, and the blue vectors indicate the shift toward the pSFH scale ($\Delta \overrightarrow{X} = \overrightarrow{X}_{\rm npSFH} - \overrightarrow{X}_{\rm pSFH}$). The average offsets in $\log M_\star$ and $\log{\rm SFR}$ between the npSFH and pSFH models are $-0.2$ and $-0.3$ dex, respectively. Blue open circles show the FRB host galaxies (“no-volume-limit” sample in Table \ref{['tab:selection']}) measured on the npSFH scale, while red circles indicate their positions after conversion to the pSFH scale. Gray arrows represent the corresponding shift vectors for individual hosts, matched to the nearest available vector in the grid.