Outline - main energy-generating mechanisms in galaxies: BH accretion and Star Formation - Starbursts and AGN’s may be linked in an evolutionary sequence - a future goal is to understand the history of the luminosity source of galaxies along galaxy evolution (e.g. SPICA & JWST) in the context of galaxy evolution - a present goal is to derive the bi-variate AGN and Star Formation luminosity functions in the Local Universe (Spitzer+Herschel) - Spitzer (and Herschel) spectroscopy trace the AGN (and its effects on the circumnuclear molecular emission, e.g. XDRs), and SF (stellar photoionization, shocks, PDRs), not suffering extinction which can obscure the galactic nuclei - IR spectroscopy (ISO & Spitzer) is able to distinguish between BH accretion and SF 1

The general context of Galaxy evolution: What we want to know 1) Full Cosmic History of Energy Generation by Stars (Fusion) and Black Holes (Accretion) (it’s not just quasars, but Seyfert galaxies which dominate at the ‘knee” of the Luminosity Function) 2) These energy production rates correspond to built up MASS (of central black hole, or galactic stars), and must--ultimately--be consistent. 3) Uncover how much of this is partly or heavily extinguished (reddening versus obscuration) 4) Seek cosmic connections between a galaxy’s stars and its massive Black Hole: understand the how and why of these systems 2

Why it has been hard to work out Optical continuum measurements alone are wholly inadequate Even optical spectroscopy on a massive scale can’t yield definitive answers Dust reddening may block our view at short wavelengths (  longer rest wavelengths) Even where dust is not a problem, redshift restricts what we can readily measure in more distant galaxies This is the ultimate multiwavelength problem, pushing instrumental sensitivities to their limits over very large areas (volumes) 3

Global Accretion power (X-rays) and Star Formation (H  ) were ~20 times higher at z= than today Recent examples of attempts to measure: Black Hole Accretion Power Young Star Power, ( but Bolometric Corrections could easily be Assuming 1 magnitude of absorption at H  off more than an order of magnitude) Hasinger et al 2005Shim et al 2009 No more H  data 4

- The main energy-generating mechanisms in galaxies are black hole accretion and star formation - Starbursts and AGN’s may be linked in an evolutionary sequence Ferrarese and Merrit 2000 (ApJ,539,L9) first publish a relation between the black-hole mass and the stellar velocity dispersion in the galactic bulges of galaxies: M BH = σ 4.8±0.5 Black hole mass Stellar velocity dispersion Shen et al 2008 (AJ,135,928) have put together much more data and conclude with: M BH = σ 3.34±0.24 The formation of spheroids (bulges) is linked to the growth of the Black Holes 5 Why it is important to isolate and measure star formation and accretion processes along galaxy evolution ?

On a cosmic scale, the evolution of supermassive black holes (SMBHs) appears tied to the evolution of the star-formation rate (SFR) (Marconi et al 2004; Merloni et al 2004). Evolution of the stellar mass density Evolution of SFR density Evolution of BH accretion rate density Merloni et al 2004 (mnras 354, L37) Heckman et al (2004) concludes a study of 23,000 low-redshift narrow emission-line AGNs of the Sloan Survey, suggesting that: the growth of black holes through Accretion and the growth of bulges through Star Formation are related at the present time in the same way that they have been related, on average, throughout cosmic history On a local scale, evidence is mounting that SF and nuclear activity are linked. Two possible evolutionary progressions are: HII  Seyfert2 (Storchi-Bergmann et al 2001; Kauffmann et al 2003), or HII  Seyfert2  Seyfert1 (Hunt & Malkan 1999; Levenson et al 2001; Krongold et al 2002). 6

FIR mid-IR near-IR optical UV 60 micron 12 micron Seyfert 2 Seyfert 1 PG quasars Spectral energy distributions of 13 AGN normalized to the bolometric fluxes (computed from µm) [Spinoglio & Malkan, 1989; Spinoglio et al. 1995] Dust absorbs the continuum at short wavelengths and re- emit it in the FIR. 7-12µm range: the absorption of the original continuum is balanced by the thermal emission. Why selecting at 12 micron ? It is the less biased local sample of active galaxies F 12µm ≈1/5 F bolometric for all types of AGN  12 µm COMPLETE SAMPLES IN BOLOMETRIC FLUX Seyfert galaxies out of a total sample of 893 galaxies (Rush, Malkan & S. 1993) What we have done in the Local Universe

8 Slope is just about 1.0, with very small scatter (also in a flux/flux diagram) Not true for 25, 60, 100µm [Spinoglio et al. 1995] Multi-wavelength energy distributions and bolometric luminosities of the 12 micron Galaxy Sample

We need to understand obscuration in active galaxies. This can be done with detailed study of the continuum (e.g. Silicate absorption), or we can use the total Hydrogen equivalent absorption column density as measured from hard X-rays. Compton thick AGN WIDE COVERAGE of the Luminosity-Hydrogen absorption column density PLANE of all objects of the 12 μ m active galaxy sample detected at hard X-rays (with a measured column density N H ). 9

What are the observables that can be linked to the major SF and AGN parameters and to OBSCURATION ? ionization density Infrared fine structure lines IR fine structure lines: - separate different physical mechanisms, - cover the ionization- density parameter space - do not suffer heavily from extinction Spinoglio & Malkan (1992) predicted for the first time the line intensities of IR lines in active and starburst galaxies, before the launch of ISO. Why infrared spectroscopy is the best tool to isolate star formation and accretion ? 10 accretion fusion

Spitzer spectroscopic survey of the 12µm Seyfert Sample – 12MSG New classification: Sy1 + Sy2 HBLR = AGN1 Sy2 non HBLR = AGN2 According to simple unification: same physical objects seen by different angles Are these an homogeneous class ? 34 Sy1 21 HBLR 20 AGN2 4 non BLR 13 non Sy In the 12MSG:Classification after optical spectropolarimetry of Tran 2001, 2003 PhD thesis of Silvia Tommasin: Tommasin+08,+09 (arXiv )

Results - Spectra HBLR z= [NeV] Obscuration Extended source strong PAH AGN2 z= Sy1 z= Compact source strong [NeV]

Game Plan: IR Spectra of ‘optical quality’ Mid-IR spectroscopy (R~600) provides a full suite of strong fine structure lines over wide range of ionization Spitzer/IRS spectra have huge SNR and good spectral resolution From Sb to Sy1: - Higher ionization lines decrease in flux and equivalent widths - PAH feature remains almost constant in flux, while its equivalent width decreases 13

Traditionally, take H recombination line, since each recombination corresponds to one photo-ionization [NeII] 12.8µm forbidden line [S III] 34µm forbidden line [Si II] 35µm forbidden line …(extended 25µm continuum) PAH features (eg 11.25µm) or LIR??? H 2 emission lines (eg µm)?? Mid-IR contains several new candidates for Star Formation Rate indicators 14

AGN/Starburst Mixing Diagrams comparing Lines and Continua Seyfert 1’s (red) and Seyfert 2’s with Polarized Broad Lines (magenta) have mid-IR emission dominated by AGN, in contrast to starbursts and LINERs (green). Seyfert 2’s without Polarized Broad Lines (cyan) are a mixed bag. Some probably have their central AGN shut down currently (i.e. this century). These ionization-sensitive []-line ratios tell the same story as the IR dust continuum: a stronger AGN contribution is closely tied to stronger (nonstellar) 12—25µm continuum from hot dust near the nucleus, with NO PAH’s. Hottest Dust Coolest Dust Strong PAH/Continuum NO PAHs 15

Equivalent width=ratio of non-AGN emission to underlying AGN mid-IR continuum Thus E.W. of non-AGN emission such as [NeII] and PAHs is inversely proportional to AGN fraction of mid-IR light. And the more AGN light, the more compact the mid-IR continuum: “Extendedness” = Ratio of 19µm continuum in big slit/small slit 16

AGN diagnostic diagrams: model Semi-analytical models to disentangle the AGN and Starburst contribution to the total galaxy emission PAH11.25µm EW [NeV]14.32µm/[NeII] [OIV]/[NeII] [NeII] EW Extendedness α (60-25)µm R = F gal 19µm /F AGN 19µm Expression of the as functions of R Theoretically for each of the galaxies there are 6 R i  if at least 3 of them are consistent with each other  reliable is computed  the AGN percentage in that galaxy is estimated. %AGN & %Sb estimated in 59 over 91 sample sources

The AGN and starburst contribution to the 19µm flux AGN + starburst at 19µm flux by inverting the analitical models: ΣR i from each of the models  R = F Sb /F AGN F Sb + F AGN = 1 %F AGN = 1/(1+R) %F Sb = R/(1+R) - Sy1: 92%±6% - HBLR: 92%±8% - AGN2: 79%±16%

[NeV] as indicator of AGN activity The [NeV] emission lines is a basic requirement for a galaxy to be classified as an AGN: It is detected in the 88% of the AGN 1's 90% of the AGN 2's 17% of the non-Sy’s Deep spectroscopic searches for [NeV] lines can discover relatively weaker AGN with lower luminosities. (cf. Goulding & Alexander 09) Sy1 non Sy

AGN statistics: Line luminosity functions Sy1 & Sy2 do not show any difference in their mid-IR line luminosity functions. New classification not used: AGN2 objects are too few to be statistically meaningful

AGN statistics: AGN power lgL AGN 19µm =0.97lgL [NeV]14µm Accretion power in the local universe z≈0.03: erg/sec [NeII] as Star Formation index, same approach  SF power in the Seyfert galaxies in the local universe: erg/sec As seyfert’s are ~10% of the total 12µm sample => total SF power ~ total accretion p. Local AGN LF(19µm) AGN fractional vs [NeV]14.3µm line luminosity

22 Main results of the Spitzer spectroscopic survey of the 12MSG new classification of Seyfert galaxies: AGN 1’s = Sy 1’s + HBLR Sy 2’s (spectropolarization), AGN 2’s = Sy 2’s without polarized broad lines. the mid-IR properties characterize AGN 1’s as an homogeneous class, while AGN 2’s have characteristics spanning from the AGN 1’s to the non-Seyferts. semi-analytic models based on the observed mid-IR spectra separate and quantify the AGN and starburst components in Seyfert galaxies. First derivation of the mid-IR line luminosity functions for Sy 1’s and 2’s and AGN 1’s and AGN 2’s, => no difference between either of the two populations. [NeV] lines = unambiguous tracers of the AGN => their luminosity functions estimate the accretion power in the local universe within a volume out to z=0.03. future work on this sample is the comparison with photoionization models.

Spitzer 9-38 μ m high resolution spectra + HERSCHEL-PACS-like spectra Tommasin, Spinoglio et al. 2008, 2009 ISO-LWS full scans, Fischer et al 1999 Obscuration extinction at Arp220 nuclei =10 4 mag i.e. N(H 2 )~10 25 cm - 2 (González-Alfonso+ 2004) Obscuration 23 Links between OBSCURATION, ACCRETION ACTIVITY and STAR FORMATION in galaxies must be determined to understand GALAXY EVOLUTION Spitzer and Herschel spectroscopy will be complementary

What is next: SPICA JAXA + ESA Cosmic Vision m telescope Cooled to < 6K Instruments cover μm -MIR spectro-photometer -FIR imaging spectrometer. -MIR Medium/High Resolution Spectrometer -MIR coronagraph -Focal Plane Camera dedicated to guidance -FIR and sub-mm spectrometer – optional

25 SPICA Sensitivity - spectroscopy Single unresolved line in single object FTS 100’s times faster to cover multiple lines over large field of view Wavelength (um) Sensitivity 5-sig 1 hour (W m -2 ) 1e-20 1e-17 1e-18 1e-19 1e-16 1e-21 ALMA JWST Spitzer ~ x15 Herschel PACS

26 Herschel and SCUBA-2  thousands of objects in photometric surveys Only spectroscopy can reveal nature and role of AGN and star formation in galaxy evolution Herschel sees local/exotic Need to detect distant objects To reveal their nature and physics and chemistry

27 Looking closer at the SPIRE background sources The Multiplex Advantage SPICA FIR FTS will take spectra of 7-10 sources/field Images Rosenbloom, Oliver, Smith, Raab private communication

Galaxies: study co-evolution of stars and black holes Redshift and nature of the sources in single shot – Evolution of the massive, dusty distant galaxy population – What is shaping the mass and luminosity functions of galaxies? – How do star formation rate and AGN activity vary with environment and cosmological epoch? 28 The first FIR spectroscopic cosmological surveys: What has been done in the local Universe with Spitzer will be done up to z~3 with SPICA:

The figure contains a prediction of the results of a spectroscopic survey over 0.5 square degrees (500 hrs with nominal sensitivity of 2x10^- 19 W/m2 1 hr, 5σ) Numbers of detected sources would be: 120 Type-1 AGNs [OIV], [NeV] 770 Type-2 AGNs “ “ 1870 starburst galaxies in [SiII] for a total of  2800 objects (Franceschini model) Gran total of  7 objects per SAFARI field 2’ x 2’ Within uncertainties two different models (Gruppioni and Franceschini) predict about 7-10 sources to be spectroscopically detected in more than 1 line down to the expected flux limits of SPICA, with about 20% of sources to be detected at z>2. Similar figures from direct integration of e.g. Magnelli et al. LF SPICA-SAFARI excellent at detecting high-z sources and at assessing in a direct way their nature (e.g whether mainly AGN or SF powered) thanks to blind spectroscopy (See Spinoglio et al.:arXiv: )

30 SPICA FIR 900 hour spectral survey Image Springel et al SPICA FIR Herschel PACS

Conclusions and future work To understand galaxy evolution we need to trace the two major energy producing processes: BH accretion and Star Formation The best tool is mid-IR and far-IR spectroscopy (both SF and AGN are often obscured by dust). At low redshift this was at reach of the current infrared space telescopes Spitzer in the mid-IR and Herschel in the far-IR. SAFARI onboard of SPICA will be able to measure AGN and starburst lines in the distant Universe. 31 Blind FIR spectroscopic surveys with SAFARI can be the mean to “physically” measure galaxy evolution The sensitivity of 2x10^-19 W/m2 (5σ, 1 hour) MUST be reached to make these surveys.