A Supernova Origin for Dust in a High-Redshift Quasar

A Supernova Origin for Dust in a High-Redshift Quasar

A SUPERNOVA ORIGIN FOR DUST IN A HIGH-REDSHIFT QUASAR R. Maiolino ∗, R. Schneider †∗, E. Oliva ‡∗, S. Bianchi §, A. Ferrara ¶, F. Mannucci§, M. Pedani‡, M. Roca Sogorb k ∗ INAF - Osservatorio Astrofisico di Arcetri, Largo Enrico Fermi 5, 50125 Firenze, Italy † “Enrico Fermi” Center, Via Panisperna 89/A, 00184 Roma, Italy ‡ Telescopio Nazionale Galileo, C. Alvarez de Abreu, 70, 38700 S.ta Cruz de La Palma, Spain § CNR-IRA, Sezione di Firenze, Largo Enrico Fermi 5, 50125 Firenze, Italy ¶ SISSA/International School for Advanced Studies, Via Beirut 4, 34100 Trieste, Italy k Astrofisico Fco. S`anchez, Universidad de La Laguna, 38206 La Laguna, Tenerife, Spain Interstellar dust plays a crucial role in the evolution of the Universe by assisting the formation of molecules1, by triggering the formation of the first low-mass stars2, and by absorbing stellar ultraviolet-optical light and subsequently re-emitting it at infrared/millimetre wavelengths. Dust is thought to be produced predominantly in the envelopes of evolved (age >1 Gyr), low-mass stars3. This picture has, however, recently been brought into question by the discovery of large masses of dust in the host galaxies of quasars4,5 at redshift z > 6, when the age of the Universe was less than 1 Gyr. Theoretical studies6,7,8, corroborated by observations of nearby supernova remnants9,10,11, have suggested that supernovae provide a fast and efficient dust formation environment in the early Universe. Here we report infrared observations of a quasar at redshift 6.2, which are used to obtain directly its dust extinction curve. We then show that such a curve is in excellent agreement with supernova dust models. This result demonstrates a supernova origin for dust in this high-redshift quasar, from which we infer that most of the dust at high redshifts has probably the same origin. arXiv:astro-ph/0409577v1 24 Sep 2004 Powerful quasars offer an optimal tool An important class of quasars are the for detailed studies of dust in their host Broad Absorption Line (BAL) quasars, galaxies out to very high redshifts. Dust whose UV spectrum is characterized by extinction is inferred through redden- blueshifted, deep and broad absorption ing of the quasar UV/optical continuum features associated with highly ionized emission. The extinction curve (Aλ = atomic species, which trace powerful out- 1.086 τλ, where τλ is the wavelength- flows of dense gas along our line of sight. dependent optical depth) of dust associ- Low Ionization BAL (LoBAL) quasars are ated with low redshift, mildly obscured characterized by additional low ionization quasars has been typically found to be absorption lines, probably associated with consistent with that of the Small Magel- higher column densities of gas. LoBAL lanic Cloud (SMC)12,13,14, while for heav- quasars at z < 4 are always significantly ily absorbed quasars there are indica- reddened by dust (associated with the out- tions that the extinction curve may be flowing gas), and therefore they are ideal different15,16. 2 R. Maiolino et al. laboratories to investigate the properties be fitted with a non-BAL (unreddened) of dust. quasar template with a slope α = −2.1 α By means of low resolution near-IR (Fλ ∝ λ ), which is shown with a dot- spectroscopic observations we have re- ted line in Fig.1. Such a blue slope is cently identified a few BAL quasars at consistent with the intrinsic slope of – z ≈ 5 − 6 (Maiolino et al. 2004). The 2.01 inferred by Reichard et al. (2003)13 most distant among them is the LoBAL for BAL quasars. The observed spec- quasar SDSSJ104845.05+463718.3 (here- trum deviates significantly from the non- after SDSS1048+46). At a redshift z = BAL unreddened template only at λrest < 6.193 (based on a new medium resolution 1700 A.˚ While such a deviation was ini- spectrum around MgIIλ2798; Maiolino et tially ascribed to a severe blend of the al. in prep), this is the only LoBAL CIV and SiIV troughs in the low resolu- at z > 5 currently known, and offers a tion spectrum, the new medium resolu- unique chance of investigating the dust tion spectrum (whose unsmoothed and en- extinction curve at z ≈ 6. This quasar larged version is shown in Fig. 1b) clearly was re-observed at higher spectral reso- shows that this is not the case. The con- lution with the goal of de-blending the tinuum outside the troughs is indeed red- deep troughs characterizing its spectrum der than expected by the extrapolation of and of estimating the reddening of the the longer wavelength spectrum through continuum. Observations were obtained the unreddened quasar template (dotted in February 2004 with the Near Infrared line). If such reddening at short wave- Camera Spectrometer (NICS18) at the lengths is due to dust, then the extinc- Telescopio Nazionale Galileo (Canary Is- tion curve must be quite unusual: rela- lands). SDSS1048+46 was observed for tively flat at λ> 1700 A˚ and steeply rising a total of 3 hours, with a slit of 1.5′′ at shorter wavelengths. We can quantita- oriented along the parallactic angle. We tively derive the extinction curve by using used a grism covering the spectral region the equation Aλ = −2.5 log(Fobs/Fintr), 0.8 − 1.45µm. where Fobs is the observed spectrum and Fig. 1a shows the observed spectrum Fintr is the intrinsic spectrum. We avoided (solid line) smoothed to a lower resolution spectral regions contaminated by emis- for sake of clarity, and combined with a sion/absorption features by interpolating previous low resolution spectrum17 of the the continuum from the nearby regions. 26 full wavelength range 0.8 − 2.4µm. At The blend of FeII lines makes more un- the quasar’s redshift our data cover the certain the description of the continuum wavelength range 1200 < λrest < 3300A˚ in the spectral region between 2300A˚ and in the rest frame of the emitted radia- 3050A;˚ however, the latter spectral re- tion. The observed spectrum is much gion is not critical within the context of bluer than LoBAL quasars observed at this paper, since the most important as- z <4, whose average spectrum is shown pect is that the extinction curve must be with the dashed line13. In particular, rather flat in the region between 1700A˚ SDSS1048+46 is bluer than any known and 3300A,˚ and increase rapidly at λrest < LoBAL quasar13 at z <4. Since the red 1700A.˚ For the intrinsic spectrum we used shape of LoBAL quasars at z<4 is ascribed the non-BAL template obtained by the 13 to dust reddening, the much bluer slope of SLOAN survey (this choice is appropri- SDSS1048 already indicates a large varia- ate since non-BAL quasars at z∼6 have 17 tion of dust extinction and reddening from spectra similar to those at z<4) , with z < 4 to z ≈ 6. slopes ranging from α = −2.1 (the red- The continuum at λrest >1700 A˚ can dest slope compatible with our observed A SUPERNOVA ORIGIN FOR DUST IN A HIGH-REDSHIFT QUASAR 3 spectrum at λrest > 1700A)˚ to α = −2.5 of silicates (Mg2SiO4 and MgSiO3), amor- (only 1% of quasars have slopes bluer than phous carbon (AC), magnetite (Fe3O4) 13 this value ). The resulting extinction and corundum (Al2O3). The typical grain curve is shown in Fig. 2 (thick solid line sizes range between 5 A˚ to 0.1 µm, with and shaded region). As expected, the ex- smaller grains being predominantly com- tinction curve inferred for this quasar at posed of silicates and magnetite (sizes ≈5– z =6.2 is quite different with respect to 15 A)˚ and larger ones by amorphous car- the SMC curve which applies to quasars bon (sizes ≈ 100 − 200 A).˚ The pre- at z < 4. The dot-dashed line in Fig. 1b dicted grain properties for different Type- indicates the non-BAL template absorbed II SNe have then been averaged over the with the extinction curve inferred by us in stellar Initial Mass Function (IMF). We Fig.2, which nicely matches the observed adopted a Salpeter IMF, but even higher- spectrum (except for the emission and ab- mass weighted Larson-like IMFs22 do not < sorption lines, whose intensity and shape change significantly the results ( ∼ 2% in depend on the physics of the ionized gas). the extinction curve). The extinction A required to match 3000A˚ The SN dust extinction properties have the observed spectrum is in the range 0.4- been derived using the standard Mie the- 0.8 mag. ory for spherical grains. Grains made The extinction curve inferred from the of AC are the main contributors to ex- most distant LoBAL can also reproduce tinction. We used the optical properties the shape of the second most distant for AC produced in an inert atmosphere BAL SDSS1044-01, a High Ionization (ACAR23), which is appropriate for the BAL (HiBAL)17 at z =5.78. The lat- SN ejecta environment. Other important ter quasar presents much lower reddening contributors to extinction were found to 24,25 than SDSS1048-46 and therefore cannot be Mg2SiO4 and Fe3O4 grains . be used to provide tight constraints on the Fig.2 shows the resulting extinction extinction curve. Nonetheless, similarly to curves, which are in a much better agree- SDSS1048-46, it is characterized a rather ment with the observations than the usual blue continuum at λrest > 1700A˚ and a SMC curve. The extinction curves were reddening at λrest < 1700A,˚ which can be obtained for various initial stellar metal- nicely fitted by using the same extinction licities. The best agreement is found for −2 −4 curve derived above and shown in Fig. 2. the Z = 10 Z⊙ and 10 Z⊙ mod- In order to interpret the observations, els, although the other models do not dif- we have computed the SN dust extinc- fer strongly and are still within the ob- tion curve by using the model proposed by servational uncertainties.

View Full Text

Details

  • File Type
    pdf
  • Upload Time
    -
  • Content Languages
    English
  • Upload User
    Anonymous/Not logged-in
  • File Pages
    7 Page
  • File Size
    -

Download

Channel Download Status
Express Download Enable

Copyright

We respect the copyrights and intellectual property rights of all users. All uploaded documents are either original works of the uploader or authorized works of the rightful owners.

  • Not to be reproduced or distributed without explicit permission.
  • Not used for commercial purposes outside of approved use cases.
  • Not used to infringe on the rights of the original creators.
  • If you believe any content infringes your copyright, please contact us immediately.

Support

For help with questions, suggestions, or problems, please contact us