2011ApJ...741L..20J

At high redshift, the universe is so young that core-collapse supernovae (SNe) are suspected to be the dominant source of dust production. However, some observations indicate that the dust production by SNe is an inefficient process, casting doubts on the existence of abundant SNe-dust in the early universe. Recently, Perley et al. reported that the afterglow of GRB 071025–an unusually red gamma-ray burst (GRB) at z ∼ 5–shows evidence for SNe-produced dust. Since this is perhaps the only high-redshift GRB exhibiting compelling evidence for SNe-dust but the result could easily be affected by small systematics in photometry, we re-examined the extinction properties of GRB 071025 using our own optical/near-infrared data at a different epoch. In addition, we tested SNe-dust models with different progenitor masses and dust destruction efficiencies to constrain the dust formation mechanisms. By searching for the best-fit model of the afterglow spectral energy distribution, we confirm the previous claim that the dust in GRB 071025 is most likely to originate from SNe. We also find that the SNe-dust model of 13 or 25 M_{x2609}without dust destruction fits the extinction property of GRB 071025 best, while pair-instability SNe models with a 170 M_{x2609} progenitor poorly fit the data. Our results indicate that, at least in some systems at high redshift, SNe with intermediate initial masses within 10-30 M_{x2609} were the main contributors for the dust enrichment, and the dust destruction effect due to reverse shock was negligible.