In pursuit of giants

The dust-to-stellar mass ratio (Mdust/M⋆) is a crucial, albeit poorly constrained, parameter for improving our understanding of the complex physical processes involved in the production of dust, metals, and stars in galaxy evolution. In this work, we explore trends of Mdust/M⋆ with different physical parameters and using observations of 300 massive dusty star-forming galaxies detected with ALMA up to z ≈ 5. Additionally, we interpret our findings with different models of dusty galaxy formation. We find that Mdust/M⋆ evolves with redshift, stellar mass, specific star formation rates, and integrated dust size, but that evolution is different for main-sequence galaxies than it is for starburst galaxies. In both galaxy populations, Mdust/M⋆ increases until z ∼ 2, followed by a roughly flat trend towards higher redshifts, suggesting efficient dust growth in the distant universe. We confirm that the inverse relation between Mdust/M⋆ and M⋆ holds up to z ≈ 5 and can be interpreted as an evolutionary transition from early to late starburst phases. We demonstrate that the Mdust/M⋆ in starbursts reflects the increase in molecular gas fraction with redshift and attains the highest values for sources with the most compact dusty star formation. State-of-the-art cosmological simulations that include self-consistent dust growth have the capacity to broadly reproduce the evolution of Mdust/M⋆ in main-sequence galaxies, but underestimating it in starbursts. The latter is found to be linked to lower gas-phase metallicities and longer dust-growth timescales relative to observations. The results of phenomenological models based on the main-sequence and starburst dichotomy as well as analytical models that include recipes for rapid metal enrichment are consistent with our observations. Therefore, our results strongly suggest that high Mdust/M⋆ is due to rapid dust grain growth in the metal-enriched interstellar medium. This work highlights the multi-fold benefits of using Mdust/M⋆ as a diagnostic tool for: (1) disentangling main-sequence and starburst galaxies up to z ∼ 5; (2) probing the evolutionary phase of massive objects; and (3) refining the treatment of the dust life cycle in simulations.