Analyses of high-redshift ultraluminous infrared (IR) galaxies traditionally use the observed optical to submillimeter spectral energy distribution (SED) and estimates of the dynamical mass as observational constraints to derive the star formation ra
te (SFR), the stellar mass, and age of these objects. An important observational constraint neglected in the analysis is the mass of dust giving rise to the IR emission. In this paper we add this constraint to the analysis of AzTEC-3. Adopting an upper limit to the mass of stars and a bolometric luminosity for this object, we construct stellar and chemical evolutionary scenarios, constrained to produce the inferred dust mass and observed luminosity before the associated stellar mass exceeds the observational limit. We find that the model with a Top Heavy IMF provided the most plausible scenario consistent with the observational constraints. In this scenario the dust formed over a period of ~200 Myr, with a SFR of ~500 Msun/yr. These values for the age and SFR in AzTEC-3 are significantly higher and lower, respectively, from those derived without the dust mass constraint. However, this scenario is not unique, and others cannot be completely ruled out because of the prevailing uncertainties in the age of the galaxy, its bolometric luminosity, and its stellar and dust masses. A robust result of our models is that all scenarios require most of the radiating dust mass to have been accreted in molecular clouds. Our new procedure highlights the importance of a multiwavelength approach, and of the use of dust evolution models in constraining the age and the star formation activity and history in galaxies.
