Evolution of the dust-to-metals ratio in high-redshift galaxies probed by GRB-DLAs

TL;DR

This study investigates the evolution of dust-to-metals ratio in high-redshift galaxies using GRB-DLAs, revealing a positive correlation with metallicity and highlighting discrepancies between different dust measurement methods.

Contribution

It provides a refined relation between dust depletion and DTM, and applies this to analyze dust origin and evolution in high-redshift galaxies using new and existing data.

Findings

DTM is generally lower than in the Milky Way.

DTM correlates positively with metallicity.

Discrepancies exist between depletion-based and extinction-based dust measurements.

Abstract

Context. Several issues regarding the nature of dust at high redshift remain unresolved: its composition, its production and growth mechanisms, and its effect on background sources. Aims. We provide a more accurate relation between dust depletion levels and dust-to-metals ratio (DTM), and to use the DTM to investigate the origin and evolution of dust in the high-redshift Universe via Gamma-ray burst damped Lyman-alpha absorbers (GRB-DLAs). Methods. We use absorption-line measured metal column densities for a total of 19 GRB-DLAs, including five new GRB afterglow spectra from VLT/X-shooter. We use the latest linear models to calculate the dust depletion strength factor in each DLA. Using these values we calculate total dust and metal column densities to determine a DTM. We explore the evolution of DTM with metallicity, and compare it to previous trends in DTM measured with different methods. Results. We find significant dust depletion in 16 of our 19 GRB-DLAs, yet 18 of the 19 have a DTM significantly lower than the Milky Way. We find that DTM is positively correlated with metallicity, which supports a dominant ISM grain-growth mode of dust formation. We find a substantial discrepancy between the dust content measured from depletion and that derived from the total V-band extinction, $A_V$ , measured by fitting the afterglow SED. We advise against using a measurement from one method to estimate that from the other until the discrepancy can be resolved.