The MOSDEF Survey: The Metallicity Dependence of X-Ray Binary Populations at z ∼ 2

Abstract

Population synthesis models predict that high-mass X-ray binary (HMXB) populations produced in low metallicity environments should be more X-ray luminous, a trend supported by studies of nearby galaxies. This trend may be responsible for the observed increase of the X-ray luminosity (L X ) per star formation rate (SFR) with redshift due to the decrease of metallicity (Z) at fixed stellar mass as a function of redshift. To test this hypothesis, we use a sample of 79 z ∼ 2 star-forming galaxies with oxygen abundance measurements from the MOSDEF survey, which obtained rest-frame optical spectra for ∼ 1500 galaxies in the CANDELS fields at 1.37 < z < 3.80. Using Chandra data from the AEGIS-X Deep, Deep Field North, and Deep Field South surveys, we stack the X-ray data at the galaxy locations in bins of redshift and Z because the galaxies are too faint to be individually detected. In agreement with previous studies, the average L X /SFR of our z ∼ 2 galaxy sample is enhanced by ≈ 0.4 − 0.8 dex relative to local HMXB L X -SFR scaling relations. Splitting our sample by Z, we find that L X /SFR and Z are anti-correlated with 97% confidence. This observed Z dependence for HMXB-dominated galaxies is consistent both with the local L X -SFR-Z relation and a subset of population synthesis models. Although the statistical significance of the observed trends is weak due to the low X-ray statistics, these results constitute the first direct evidence connecting the redshift evolution of L X /SFR and the Z dependence of HMXBs.

X-Ray Binary Luminosity Function Scaling Relations in Elliptical Galaxies: Evidence for Globular Cluster Seeding of Low-mass X-Ray Binaries in Galactic Fields

“…These stellar masses are based on photometry from the areal regions defined in Columns (4)-( 5) and Columns (7)-( 8), excluding a central 3″ circular region and any sky coverage that does not have HST exposure (see Section 3.1 for details). The cumulative stellar mass of the sample is have different star formation histories, GC S N values, and metallicities that can have an effect on the XRB populations (see, e.g., Fragos et al 2008Fragos et al , 2013aFragos et al , 2013bBasu-Zych et al 2013Lehmer et al 2014Lehmer et al , 2019Brorby et al 2016;Fornasini et al 2019).…”

X-ray properties of He ii λ 1640 emitting galaxies in VANDELS

“…As shown by Fornasini et al (2019), the above mentioned approximation to stacking works for galaxies that have similar L X . Since in this study we are probing galaxies with similar physical properties such as redshifts, SFRs and masses, and the luminosities that go into the stack are weighted by the effective exposure time, we do not expect large inaccuracies in the stacked luminosity measured in this way.…”

“…Within the circular aperture and the annulus, we measure the total number of counts from the source and the background, C gal and C bkg , the area encompassed, A gal and A bkg (pixel 2 ), and the effective exposure times, t gal and t bkg (s), respectively. We follow Fornasini et al (2019) and calculate the background subtracted counts as…”

“…To place our measurements of L X /SFR for both He II emitters and non-emitters within the general population of star-forming galaxies at high redshifts, we compare our measurements to those in the literature. We begin by looking at the redshift evolution of L X /SFR inferred from samples of star-forming galaxies at z ∼ 2 from Fornasini et al (2019). In Fig.…”

“…Observations of Xray emission from star-forming galaxies[at fixed star formation rates (SFRs)] have revealed a strong metallicity dependence of their X-ray luminosities. This means that the contribution from XRBs increases with decreasing metallicities (Basu-Zych et al 2013b;Douna et al 2015;Brorby et al 2016;Lehmer et al 2016;Fornasini et al 2019).…”