Résumé
Two distinct scenarios for the origin of the similar to 4 x 10(8) M-circle dot of dust observed in the high-redshift (z = 6.4) quasar J1148+5251 have been proposed. The first assumes that this galaxy is much younger than the age of the universe at that epoch so that only supernovae (SNe) could have produced this dust. The second scenario assumes a significantly older galactic age, so that the dust could have formed in lower-mass asymptotic giant branch (AGB) stars. Presenting new integral solutions for the chemical evolution of metals and dust in galaxies, we offer a critical evaluation of these two scenarios and observational consequences that can discriminate between the two. We show that AGB stars can produce the inferred mass of dust in this object, however, the final mass of surviving dust depends on the galaxy's star formation history (SFH). In general, SNe cannot produce the observed amount of dust unless the average SN event creates over similar to 2 M-circle dot of dust in its ejecta. However, special SFHs can be constructed in which SNe can produce the inferred dust mass with a reasonable average dust yield of similar to 0.15 M-circle dot. The two scenarios propose different origins for the galaxy's spectral energy distribution, different star formation efficiencies and stellar masses, and consequently different comoving number densities of J1148+5251-type hyperluminous infrared (IR) objects. The detection of diagnostic mid-IR fine structure lines and more complete surveys determining the comoving number density of these objects can discriminate between the two scenarios.