Time Delays Between the Onset of Star Formation and Nuclear Activity in Quasars Gabriela Canalizo, Mark Hancock, Mariana Lazarova, Kyle Hiner (UCR), Vardha Bennert (UCSB), Alan Stockton (IfA) Recent observations of low redshift quasar host galaxies indicate that mergers and significant episodes of star formation are ubiquitous in these galaxies. By constraining the timescales of such events we can gain a better understanding of the role of AGN feedback in galaxy evolution. We discuss results from a long campaign of space- and ground-based imaging and spectroscopic observations of z < 0.5 hosts that imply that mergers are indeed essential for the triggering of quasars, and that these mergers invariably induce starbursts either during or shortly after the merger. There appears to be, however, a large range of values for the time delays between the merger and the onset of the nuclear activity, varying from a few million years to more than a billion years. We find some evidence for a bimodal distribution, although this could be a selection effect. Abstract Quasars in Young Mergers Quasars in Older (~Gyr) Mergers Recent mergers and star formation play a prominent role in at least one fourth of the host galaxies in the quasar population at low-z (Canalizo et al. 2006,NewAR, 50, 650), and this fraction could be as high as 80% if “red quasars” are truly as abundant as IR surveys suggest (e.g., Lacy et al. 2002, AJ, 123, 2925). Most of these quasar hosts show overt signs of tidal interaction and relatively young post-starburst populations of a few hundred Myr (e.g., Canalizo et al. 2001). However, only a minority show current star formation, and the inferred SFRs are less than a few M  yr -1 (Ho 2005, ApJ, 629, 680). Acknowledgements This work was supported in part under proposals GO-10421, GO-11101 and GO-11157 by NASA through a grant from the Space Telescope Science Institute, which is operated by the Association of Universities for Research in Astronomy, Inc., under contract NAS5-26555 and under program 50792 of the Spitzer Space Telescope, which is operated by the JPL, Caltech, under a contract with NASA. Additional support was provided by the National Science Foundation, under grant number AST 0507450. Fig.1 (above): Ground-based and HST images of a sample of FIR-loud quasar host galaxies. Most show overt signs of tidal interaction with young dynamical ages as given in the table below. Fig. 2 (below): Representative post-starburst spectra of quasar hosts from the same sample. Bruzual & & Charlot ‘03 Mara aston ‘05 Object Starburst Age (Gyr) Contribution by Mass Starburst Age (Gyr) Contribution by Mass 0054+144 2.1 30% 2.0 60% 0137+012 1.7 50% 2.0 100% 0244+194 1.4 100% 1.5 100% 0736+017 2.2 10% 2.0 14% 0923+201 1.0 20% 0.6 30% 1004+130 1.0 18% 0.9 20% 1012+008 0.2 0.8% 0.6 4% 1020-103 1.5 30% (0.1) (0.6%) 1217+023 0.6 3.0% 0.6 1.5% 1635+119 1.4 52% 1.0 45% 2135-147 2.4 90% 3.0 100% 2141+175 0.7 9.1% 0.6 9.3% 2247+140 --- 0% 1.0 0.8% 2349-014 1.0 7.2% 0.8 5.9% 0054+144 0923+201 Object Starburst Age (Myr) Dynamical Age (Myr) IR00275-2859 50 130 IR04505-2958 130 50 PG 1543+489 < 200 (80) Mrk 231 42 110 IR07598+6508 32 160 PG 1700+518 85 40 3C 48 9 200 Mrk 1014 240 330 I Zw 1 (4) … A large fraction of optically-selected quasars at low-z appear to reside in the centers of galaxies that have relaxed light distributions like ellipticals. We are conducting a coordinated study with HST imaging and Keck LRIS spectroscopy of a sample of 19 z~0.2 bulge-dominated quasar host galaxies to search for fine structure indicative of past merger events as well signs of starbursts that may have been triggered by these mergers. Indeed, we are finding that the majority of objects show traces of major starburst episodes of intermediate age (typically 1-2 Gyr) involving a substantial fraction of the mass of the stellar component. In addition, the observed tidal structures are of the order of a Gyr (a few orbital times), indicating that a merger was likely responsible for the star forming episode (Canalizo et al. 2007, ApJ, 669, 801; Bennert et al. 2008, ApJ, 677, 846; Hancock et al., in preparation). Fig.3 (left). Deep HST/WFPC2 and ACS images of 15 of the 19 bulge-dominated quasar host galaxies in our sample. For the galaxies shown in color, the PSF and an elliptical galaxy model have been subtracted using GALFIT to show more clearly the signs of tidal interaction. The two galaxies in gray-scale are displayed at high contrast to show tidal tails extending to more than 50 kpc. Fig.4 (right): Keck LRIS spectra of eight of the quasar host galaxies in our sample. Integration times are 1-3 hours and each spectrum has been corrected for quasar contamination. The red trace is the best fit model typically consisting of a 10 Gyr population and an intermediate age starburst as indicated in the table below. Caught in the Act? High SFRs in LoBAL Quasars Our observations of low-redshift quasar host galaxies show that these objects tend to fall in one of two groups: 1) recent or on-going mergers with very young (a few hundred Myr) post-starburst populations, or 2) evolved merger remnants with intermediate-age stellar populations. The second group is particularly intriguing: If the observed ~Gyr-old merger in each object triggered both star formation and AGN activity, why do we still observe the AGN activity at such a late stage? It is possible that the merger that occurred over one Gyr ago may have triggered a first episode of accretion onto the black hole(s). Feedback from the quasar would have quenched any further star formation. Assuming theoretical estimates for the duration of quasar activity are correct (e.g., 10 7 -10 8 yr; Yu & Tremaine 2002, MNRAS 335, 96), the AGN activity would have ceased as the merger continued its course and the morphology of the newly merged galaxies began to relax into the shape of an elliptical. Eventually, the extended tidal debris would “rain” back into the central regions of the galaxy, triggering a new episode of quasar activity. A time delay in the onset of quasar activity would be in agreement with predictions by hydrodynamical simulations of merging galaxies. The rare subclass of optically-selected quasars known as low-ionization broad absorption line (LoBAL) quasars show signs of high-velocity gas outflows and reddened continua indicative of dust obscuration. Recent studies suggest that LoBAL quasars represent a short-lived phase early in the life of quasars, when powerful AGN-driven winds are blowing away the dust and gas surrounding the quasar. If so, these objects may be in a unique and brief phase when ongoing star formation and nuclear activity coexist. We are conducting a systematic multiwavelength study of a volume limited sample of LoBAL quasars at 0.5 < z < 0.6 drawn from SDSS. We obtained Spitzer MIPS and IRS observations to measure their star formation rates and are currently obtaining HST WFC3 images to study their morphologies. In contrast to typical low-z quasars, we find very high SFRs in loBAL quasars. Using Kennicutt’s relation (1998) we find an average of ~150 M  yr -1 for the sample, with a range from 40 to 630 M  yr -1 (Lazarova et al., in preparation). We note, however, that these figures are likely to be overestimated since the quasar continuum contributes significantly to the far-IR luminosities; in addition, we have found in related studies of quasars that even some of the cool far-IR emitting dust may be heated by AGN rather than by stellar emission (Hiner et al. 2009, ApJ, 706, 508). We must therefore model the SEDs of loBAL quasars including all contributing components to obtain reliable SFRs. Discussion