THE LUMINOUS POLYCYCLIC AROMATIC HYDROCARBON EMISSION FEATURES: APPLICATIONS TO HIGH REDSHIFT GALAXIES AND ACTIVE GALACTIC NUCLEI A Dissertation by HEATH VERNON SHIPLEY Submitted to the Office of Graduate and Professional Studies of Texas A&M University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Chair of Committee, Casey J. Papovich Committee Members, Darren DePoy Bashkar Dutta Ping Yang Head of Department, Lewis A. Ford August 2015 Major Subject: Physics & Astronomy Copyright 2015 Heath Vernon Shipley ABSTRACT Theco-evolutionofstar-formationandsupermassiveblackhole(SMBH)accretion in galaxies is one of the key problems in galaxy formation theory. Understanding the formation of galaxies, and their subsequent evolution, will be coupled to intensive study of the evolution of SMBHs. This thesis focuses on studying diagnostics of star-formation and SMBH accretion to develop tools to study this co-evolution. Chapter 2 consists of using mid-infrared (mid-IR) spectroscopy from the Spitzer Infrared Spectrograph (IRS) to study the nature of star-formation and SMBH ac- cretion. The mid-IR spectra cover wavelengths 5-38µm, spanning the polycyclic aromatic hydrocarbon (PAH) features and important atomic diagnostic lines. We divide our sample into a subsample of galaxies with Spitzer IRAC colors indicative of warm dust heated by an AGN (IRAGN) and those galaxies whose colors indi- cate star-formation processes (non-IRAGN). In both the IRAGN and star-forming samples, the luminosity in the PAH features correlates strongly with [Ne II]λ12.8µm emissionline, fromwhichweconcludethatthePAHluminositydirectlytracesthein- stantaneousstar-formationrate(SFR)inboththeIRAGNandstar-forminggalaxies. ThereisnomeasurabledifferencebetweenthePAHluminosityratiosofL /L and 11.3 7.7 L /L for the IRAGN and non-IRAGN, suggesting that AGN do not significantly 6.2 7.7 excite or destroy PAH molecules on galaxy-wide scales. In chapter 3, I calibrate the PAH luminosity as a SFR indicator. We provide a new robust SFR calibration using the luminosity emitted from PAH molecules at 6.2µm, 7.7µm and 11.3µm. The PAH features emit strongly in the mid-IR mitigating dust extinction, containing on average 5−10% of the total IR luminosity in galaxies. We use mid-IR spectroscopy from the Spitzer/IRS, and data covering other SFR ii indicators (Hα emission and rest-frame 24µm continuum emission). The PAH lumi- nosity correlates linearly with the SFR as measured by the Hα luminosity (corrected for attenuation using the mono-chromatic rest-frame 24um emission), with a tight scatter of <0.15 dex. The scatter is comparable to that between SFRs derived from the Paα and dust-corrected Hα emission lines, implying the PAH features may be as accurate a SFR indicator as the Hydrogen recombination lines. Because the PAH features are so bright, our PAH SFR calibration enables an efficient way to measure SFRs in distant galaxies with JWST to SFRs as low as ∼10 M(cid:12) yr−1 to z ∼< 2. We use Spitzer/IRS observations of PAH features in lensed star-forming galaxies at 1 < z < 3 to demonstrate the utility of the PAHs to derive SFRs as accurate as those available from Paα. Chapter 4 is the application of the PAH SFRs for galaxies with AGN to demon- strate the reliability for studies of the co-evolution of star-formation and SMBH accretion. We present a study of the contribution from star-formation in galaxies of varying AGN activity (from pure star-forming galaxies to quasars) as a function of total IR luminosity using a sample of 220 galaxies. We use mid-IR spectroscopy fromtheSpitzer/IRSandphotometryfromtheMIPS24µm, 70µmand160µmbands with partial coverage of the sample with the Herschel 160µm band for the quasars. The contribution from star-formation to the total IR luminosity implied by the PAH emission decreases with increasing IR luminosity. We find a similar result to previous studies for the correlation between SFR, i.e. PAH luminosity, and AGN luminosity for quasars of L ∝ L0.67±0.10 and L ∝ L0.55±0.15 for the 11.3µm PAH feature SF AGN SF AGN only (which has been shown to be the most reliable PAH feature in the vicinity of AGN). This may indicate the PAH luminosity remains a reliable tracer of the SFR for galaxies with strong AGN contributions (i.e. quasars), as we did not subtract off the AGN component before measuring the SFR from the PAH luminosity. iii ACKNOWLEDGEMENTS We thank our collaborators that contributed valuable insight and comments to improve the thesis: George Rieke (University of Arizona, Arizona), John Moustakas (Siena College, NY), Michael Brown (Monash University, Australia), Arjun Dey (NOAO, AZ), Buell Jannuzi (University of Arizona and NOAO, AZ) and Benjamin Weiner (University of Arizona, AZ). A very special thank you to my advisor Casey Papovich who helped guide me throughout my thesis and provided invaluable advice to me all along the way. We thank Nicholas Suntzeff and Darren DePoy for comments that helped im- prove the thesis. We also thank Rob Kennicutt Jr. and Daniela Calzetti for valuable discussions that added to the improvement of the thesis. We thank our colleagues on the NDWFS, AGES teams. We thank JD Smith for comments that helped improve the thesis. This work utilized the PAHFIT IDL tool for decomposing IRS spec- tra, which J. D. Smith has generously made publicly available (Smith et al., 2007). We thank the MPA/JHU collaboration for SDSS studies for making their catalogs publicly available. Support for the work presented in this thesis was provided by the NASA Astro- physicsDataAnalysisProgram(ADAP)fortheproposal“MeasuringStar-Formation Rates of AGNs and QSOs using a new calibration from Polycyclic Aromatic Hydro- carbon Emission” (14-ADAP14-0228), submitted in response to NNH14ZDA001N- ADAP, Research Opportunities in Space and Earth Sciences (ROSES-2014). Further support for this work was provided to the author by the George P. and Cynthia Woods Mitchell Institute for Fundamental Physics & Astronomy. This work made use of images and/or data products provided by the NOAO iv Deep Wide-Field Survey (Jannuzi & Dey, 1999; Jannuzi et al., 2004; Dey et al., 2004), which is supported by the National Optical Astronomy Observatory (NOAO). NOAO is operated by AURA, Inc., under a cooperative agreement with the National ScienceFoundation. TheresearchofADissupportedbyNOAO,whichisoperatedby the Association of Universities for Research in Astronomy, Inc., under a cooperative agreementwiththeNSF.Thisworkisbasedinpartonobservationsandarchivaldata obtained with the Spitzer Space Telescope, which is operated by the Jet Propulsion Laboratory, California Institute of Technology under a contract with NASA. Partial support for this work was provided by NASA through awards 1255094 and 1365085 issued by JPL/Caltech. Funding for the SDSS has been provided by the Alfred P. Sloan Foundation, the Participating Institutions, NASA, NSF, the U.S. Department of Energy, the Japanese Monbuk- agakusho, the Max Planck Society, and the Higher Educa- tion Funding Council for England. The SDSS Web site is http://www.sdss.org/. The SDSS is managed by the Astrophysical Research Consortium for the Participating Institutions. The Participating Institutions are the American Museum of Natural History, Astrophysical Institute Potsdam, University of Basel, University of Cam- bridge, Case Western Reserve University, University of Chicago, Drexel University, Fermilab, the Insti- tute for Advanced Study, the Japan Participation Group, Johns Hopkins University, the Joint Institute for Nuclear Astrophysics, the Kavli Institute for Particle Astrophysics and Cosmology, the Korean Scientist Group, the Chinese AcademyofSciences(LAMOST),LosAlamosNationalLaboratory, theMax-Planck- Institute for Astronomy (MPIA), the Max-Planck-Institute for Astrophysics (MPA), New Mexico State University, Ohio State University, University of Pittsburgh, Uni- versity of Portsmouth, Princeton University, the United States Naval Observatory, and the University of Washington. v TABLE OF CONTENTS Page ABSTRACT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ii ACKNOWLEDGEMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . iv TABLE OF CONTENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vi LIST OF FIGURES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix LIST OF TABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xix 1. INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 2. DIAGNOSTICS OF AGN AND STAR-FORMATION AND THE IMPOR- TANCE OF PAH EMISSION USING SPITZER SPECTROSCOPY OF INFRARED-LUMINOUS GALAXIES . . . . . . . . . . . . . . . . . . . . 4 2.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 2.2 Sample Definition and Selection . . . . . . . . . . . . . . . . . . . . . 7 2.3 Spitzer Imaging Data . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 2.3.1 Bo¨otes Field . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 2.3.2 Spitzer First Look Survey (FLS) . . . . . . . . . . . . . . . . 15 2.3.3 IRAC AGN Selection . . . . . . . . . . . . . . . . . . . . . . . 15 2.4 IRS Observations, Data Reduction, and Analysis . . . . . . . . . . . . 18 2.4.1 IRS Observations . . . . . . . . . . . . . . . . . . . . . . . . . 18 2.4.2 Data Reduction . . . . . . . . . . . . . . . . . . . . . . . . . . 18 2.4.3 IRS Spectral Fitting . . . . . . . . . . . . . . . . . . . . . . . 20 2.4.4 Offset Between IRS SL and LL Modules . . . . . . . . . . . . 25 2.5 Comparison of Mid-IR Emission Features and Relation to Total In- frared Luminosity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 2.5.1 Composite Spectra . . . . . . . . . . . . . . . . . . . . . . . . 26 2.5.2 Measuring the Total Infrared Luminosity . . . . . . . . . . . . 28 2.5.3 Contribution of PAH Emission to L . . . . . . . . . . . . . . 29 IR 2.5.4 Detection Frequency of Emission Features . . . . . . . . . . . 31 2.5.5 Measures of Grain Sizes and Ionization State . . . . . . . . . . 35 2.5.6 Relation between Radiation Hardness and PAH Strength . . . 37 2.5.7 The Distribution of the 6.2µm PAH Equivalent Width . . . . 40 vi 2.5.8 The Relationship Between PAH Luminosity, Star-formation Rate, and AGN Luminosity . . . . . . . . . . . . . . . . . . . 43 2.6 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 2.6.1 The Color-Magnitude Diagram for IR Luminous Galaxies . . . 44 2.6.2 AGN Effects on PAH Emission and AGN Contribution to L 49 IR 2.6.3 EmissionRatiosofShort-to-LongWavelengthPAHsinIRAGN and non-IRAGN . . . . . . . . . . . . . . . . . . . . . . . . . 52 2.6.4 Galaxies with Excess [O IV] λ25.9µm Emission . . . . . . . . 54 3. ANEWSTAR-FORMATIONRATECALIBRATIONFROMPOLYCYCLIC AROMATICHYDROCARBONEMISSIONFEATURES:APPLICATION TO HIGH REDSHIFT GALAXIES . . . . . . . . . . . . . . . . . . . . . . 58 3.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 3.2 Sample and Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 3.2.1 Calibration Samples . . . . . . . . . . . . . . . . . . . . . . . 61 3.2.2 Application Sample of High Redshift Galaxies . . . . . . . . . 65 3.3 Derived Quantities . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 3.3.1 IRS Spectral Fitting . . . . . . . . . . . . . . . . . . . . . . . 68 3.3.2 Optical Spectral Fitting . . . . . . . . . . . . . . . . . . . . . 69 3.3.3 Dust-Corrected Hα Emission from Rest-Frame MIPS 24 µm Flux . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 3.3.4 Gas-Phase Metallicities . . . . . . . . . . . . . . . . . . . . . . 73 3.3.5 Total Infrared Luminosity . . . . . . . . . . . . . . . . . . . . 75 3.4 The PAH SFR Calibration . . . . . . . . . . . . . . . . . . . . . . . . 77 3.4.1 PAH SFR Relations . . . . . . . . . . . . . . . . . . . . . . . 77 3.4.2 Uncertainties for Derived SFR Relations . . . . . . . . . . . . 84 3.4.3 Correction to PAH Luminosity for Low Metallicity Galaxies . 86 3.4.4 The Interesting Galaxy: II Zw 096 . . . . . . . . . . . . . . . 87 3.5 Comparison to Previous Studies of PAH SFR Calibrations . . . . . . 90 3.5.1 Comparison to PAH SFRs Calibrated to the Total Infrared Luminosity . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92 3.5.2 Comparison to PAH SFRs Calibrated to mid-IR Atomic Emis- sion Lines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 3.6 SFRs of High Redshift (Lensed) Galaxies: A Preview for JWST . . . 97 3.6.1 SFRs for Galaxies at z ∼ 1 . . . . . . . . . . . . . . . . . . . . 101 3.6.2 SFRs for Galaxies at 2 < z < 3 . . . . . . . . . . . . . . . . . 101 4. THEROLEOFSTAR-FORMATIONINGALAXIESFROMSTARBURSTS TO QUASARS AS A FUNCTION OF TOTAL IR LUMINOSITY USING PAH EMISSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 4.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 vii 4.2 Sample and Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108 4.2.1 O’Dowd et al. (2009) Sample . . . . . . . . . . . . . . . . . . 110 4.2.2 Shipley et al. (2013) Sample . . . . . . . . . . . . . . . . . . . 110 4.2.3 Shi et al. (2014) Sample . . . . . . . . . . . . . . . . . . . . . 111 4.3 Analysis of PG and 2MASS Quasars for Shi et al. (2014) Sample . . . 112 4.3.1 Spitzer IRS Spectral Fits . . . . . . . . . . . . . . . . . . . . . 112 4.3.2 Total IR Luminosity . . . . . . . . . . . . . . . . . . . . . . . 114 4.4 Distribution Functions of Star-Formation and AGN Luminosities . . . 115 4.4.1 Star-Formation as a Function of Total IR Luminosity . . . . . 115 4.4.2 Star-Formation in Quasars . . . . . . . . . . . . . . . . . . . . 120 4.5 Disscussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122 4.5.1 The Correlation Between L and L . . . . . . . . . . . . . 122 SF AGN 4.5.2 Comparison to Models . . . . . . . . . . . . . . . . . . . . . . 125 5. SUMMARY AND CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . 127 REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135 APPENDIX A. ESTIMATING THE TOTAL IR LUMINOSITY . . . . . . . 155 viii LIST OF FIGURES FIGURE Page 2.1 — Color-Magnitude Diagram using optical data from the AGES cata- logforoursample. Thegrey-shadingindicatesthedensityofallgalax- iesfromAGESinthatregionofcolor-magnitudespaceand0.02 < z < 0.6, where the grey-shading increases as the density increases linearly. The cyan points show those AGES sources with 0.02 < z < 0.6 and f(24µm) ≥ 1.2 mJy, the IRS spectroscopic limit for our sample. The galaxies selected in our sample are indicated by white squares, white circles, and yellow diamonds for Dole (program 20113), Lagache (pro- gram 20128, non-FLS sources), and Rieke (program 40251), respec- tively. Furthermore, we selected IR-luminous galaxies for our sample such that they span the full range of (u−r) optical color with an 0.1 equal number (12-13) galaxies in each of four bins, denoted by the red-dashed lines and defined in section 2.2. . . . . . . . . . . . . . . . 10 2.2 —Distributionoftheredshiftsand24µmfluxdensitiesofthe65galax- ies in our IRS sample. The galaxies are indicated by white squares, white circles, and yellow diamonds for Dole (program 20113), Lagache (program 20128), and Rieke (program 40251), respectively. The or- ange circles are from O’Dowd et al. (2009) sample as a comparison to our sample. The red dashed line indicates f (24µm) = 1.2mJy. The ν redshiftmeanandmedianofthedistributionare0.30and0.28, respec- tively. The redshift distribution is fairly uniform from 0.2 < z < 0.6, with an interquartile range (which contains the inner 50% of galaxies) of z = 0.18−0.42. . . . . . . . . . . . . . . . . . . . . . . . 11 interquartile 2.3 — IRAC colors of [5.8] - [8.0] versus [3.6] - [4.5] for galaxies in our IRS sample. Here, the IRAC colors are in the Vega magnitude system, fol- lowingSternetal.(2005). ThebluecirclesaresourcesfromtheBoo¨tes field and green diamonds are FLS sources. The dotted lines show the empirical selection-criteria for IRAC-AGN selection: galaxies inside this “wedge” are IRAC-selected AGN (Stern et al., 2005). We denote the subsample of galaxies in our IRS sample that satisfy these colors as IRAGN. We denote galaxies outside this wedge as non-IRAGN, and we expect the IR emission in these objects to be dominated by star-formation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 ix 2.4 — Examples of PAHFIT spectral decomposition to object 23 (top) and object 4 (bottom). In each panel, the IRS spectroscopic data are shown as black squares. The total fit and individual spectral components fit by PAHFIT are shown, including the molecular and atomicemissionfeatures(bluecurves), dustcontinua(red), andstellar light (magenta). The total fit (the sum of all the model components) is shown in green, which provides a good representation of the data. . 20 2.5 — Composite spectra for the IRS sample. The top panel shows the composite spectrum for all 65 galaxies in the IRS sample. The mid- dle and bottom panels show composite spectra for the subsamples of IRAGN (14 galaxies) and non-IRAGN (51 galaxies), respectively. The vertical red-dashed lines indicate the prominent PAH features and emission lines in the wavelength range, as labeled. The flux is normalized at the continuum flux of 21µm. The outset panels show the composite spectra in a small wavelength region 24.5µm-27.5µm to show the strength of the [O IV] λ25.9µm emission line (see sec- tion 2.5.1 for explanation). The error bars shown is the error on the weighted mean for each composite spectrum. . . . . . . . . . . . . . . 27 2.6 — Redshift versus the total IR luminosity from 8-1000µm for our IRS sample (blue circles, Boo¨tes sources; green diamonds, FLS sources) derived from the MIPS 24 µm data (and 70 and 160 µm, if available) derived using the Rieke et al. (2009) IR SEDs. The galaxies in our IRSsamplespantherangeofIRluminosityof“LuminousIRgalaxies” (LIRGs), L =1011−1012L . ThefigurealsoshowstheO’Dowdetal. IR (cid:12) (2009) SSGSS sample (orange circles), which are lower redshift and IR luminosity. The dashed curve shows the limiting IR luminosity as a function of redshift for a fixed 24 µm flux density of 1.2 mJy using the IR SEDs from Rieke et al. (2009). . . . . . . . . . . . . . . . . . . . . 30 2.7 — The distribution of L /L for our sample, where L is the PAH IR PAH total luminosity of the 6.2, 7.7, 8.6, 11.3, 12.7, and 17.0µm PAH emis- sion features. The bottom panel shows the distribution for our IRS sample. The middle panel shows the distribution for our subsample of IRAGN, and the top panel shows the distribution for our subsample of non-IRAGN. The median L /L is 0.05 for the IRAGN, and is PAH IR about half that for the non-IRAGN which have a median of L /L PAH IR = 0.09. The arrows represent the L /L ratios from the composite PAH IR spectra in figure 2.5 (L /L = 0.09, 0.06, and 0.10 for the IRS PAH IR sample, IRAGN, and non-IRAGN, respectively). . . . . . . . . . . . . 32 x
Description: