Explaining the [CII]158um Deficit in Luminous Infrared Galaxies - First Results from a Herschel/PACS Study of the GOALS Sample
T. Diaz-Santos, L. Armus, V. Charmandaris, S. Stierwalt, E. J. Murphy, S. Haan, H. Inami, S. Malhotra, R. Meijerink, G. Stacey, A. O. Petric, A. S. Evans, S. Veilleux, P. P. van der Werf, S. Lord, N. Lu, J. H. Howell, P. Appleton, J. M. Mazzarella, J. A. Surace, C. K. Xu, B. Schulz, D. B. Sanders, C. Bridge, B. H. P. Chan, D. T. Frayer, K. Iwasawa, J. Melbourne, E. Sturm
aa r X i v : . [ a s t r o - ph . C O ] J u l Draft version September 13, 2018
Preprint typeset using L A TEX style emulateapj v. 5/2/11
EXPLAINING THE [C ii ]157.7 µ m DEFICIT IN LUMINOUS INFRARED GALAXIES – FIRST RESULTS FROMA HERSCHEL/PACS STUDY OF THE GOALS SAMPLE T. D´ıaz-Santos † , L. Armus , V. Charmandaris , S. Stierwalt , E. J. Murphy , S. Haan , H. Inami S. Malhotra , R. Meijerink , G. Stacey , A. O. Petric , A. S. Evans , S. Veilleux , P. P. van der Werf ,S. Lord , N. Lu , J. H. Howell , P. Appleton , J. M. Mazzarella , J. A. Surace , C. K. Xu B. Schulz
D. B. Sanders , C. Bridge , B. H. P. Chan , D. T. Frayer , K. Iwasawa , J. Melbourne , and E. Sturm Spitzer Science Center, California Institute of Technology, MS 220-6, Pasadena, 91125, CA IESL/Foundation for Research and Technology - Hellas, GR-71110, Heraklion, Greece and Chercheur Associ´e, Observatoire de Paris,F-75014, Paris, France University of Crete, Department of Physics, GR-71003, Heraklion Department of Astronomy, University of Virginia, P.O. Box 400325, Charlottesville, VA 22904 Observatories of the Carnegie Institution for Science, 813 Santa Barbara Street, Pasadena, CA 91101, USA CSIRO Astronomy and Space Science, Marsfield NSW, 2122, Australia National Optical Astronomy Observatory, 950 N. Cherry Ave., Tucson, AZ 85719, USA School of Earth and Space Exploration, Arizona State University, Tempe, AZ 85287, USA Kapteyn Astronomical Institute, University of Groningen, P.O. Box 800, NL-9700 AV Groningen, The Netherlands Department of Astronomy, Cornell University, Ithaca, NY 14853, USA Astronomy Department, California Institute of Technology, Pasadena, CA 91125, USA National Radio Astronomy Observatory, 520 Edgemont Road, Charlottesville, VA 22903 Joint Space-Science Institute, University of Maryland, College Park, MD 20742, USA Department of Astronomy, University of Maryland, College Park, MD 20742, USA Leiden Observatory, Leiden University, P.O. Box 9513, NL-2300 RA Leiden, The Netherlands NASA Herschel Science Center, IPAC, California Institute of Technology, MS 100-22, Cech, Pasadena, CA 91125 Infrared Processing and Analysis Center, MS 100-22, California Institute of Technology, Pasadena, CA 91125 Institute for Astronomy, University of Hawaii, 2680 Woodlawn Drive, Honolulu, HI 96822 National Radio Astronomy Observatory, P.O. Box 2, Green Bank, WV 24944, USA ICREA and Institut de Cincies del Cosmos (ICC), Universitat de Barcelona (IEEC-UB), Marti i Franques 1, 08028 Barcelona, Spain Caltech Optical Observatories, Division of Physics, Mathematics and Astronomy, MS 301-17, California Institute of Technology,Pasadena, CA 91125, USA and Max-Planck-Institut f¨ur extraterrestrische Physik, Postfach 1312, D-85741 Garching, Germany
Draft version September 13, 2018
ABSTRACTWe present the first results of a survey of the [C ii ]157.7 µ m emission line in 241 luminous infraredgalaxies (LIRGs) comprising the Great Observatories All-sky Survey (GOALS) sample, obtained withthe PACS instrument on board the Herschel Space Observatory . The [C ii ] luminosities, L [C II] , ofthe LIRGs in GOALS range from ∼ to 2 × L ⊙ . We find that LIRGs show a tight correlationof [C ii ]/FIR with far-IR flux density ratios, with a strong negative trend spanning from ∼ − to 10 − , as the average temperature of dust increases. We find correlations between the [C ii ]/FIRratio and the strength of the 9.7 µ m silicate absorption feature as well as with the luminosity surfacedensity of the mid-IR emitting region (Σ MIR ), suggesting that warmer, more compact starbursts havesubstantially smaller [C ii ]/FIR ratios. Pure star-forming LIRGs have a mean [C ii ]/FIR ∼ × − ,while galaxies with low 6.2 µ m PAH equivalent widths (EWs), indicative of the presence of activegalactic nuclei (AGN), span the full range in [C ii ]/FIR. However, we show that even when only purestar-forming galaxies are considered, the [C ii ]/FIR ratio still drops by an order of magnitude, from10 − to 10 − , with Σ MIR and Σ IR , implying that the [C ii ]157.7 µ m luminosity is not a good indicatorof the star formation rate (SFR) for most LIRGs, for it does not scale linearly with the warm dustemission most likely associated to the youngest stars. Moreover, even in LIRGs in which we detect anAGN in the mid-IR, the majority (2/3) of galaxies show [C ii ]/FIR ≥ − typical of high 6.2 µ m PAHEW sources, suggesting that most AGNs do not contribute significantly to the far-IR emission. Weprovide an empirical relation between the [C ii ]/FIR and the specific SFR (SSFR) for star-formingLIRGs. Finally, we present predictions for the starburst size based on the observed [C ii ] and far-IRluminosities which should be useful for comparing with results from future surveys of high-redshiftgalaxies with ALMA and CCAT. Subject headings: galaxies: nuclei — galaxies: starburst — galaxies: ISM — infrared: galaxies INTRODUCTION
Systematic spectroscopic observations of far-infrared(IR) cooling lines in large samples of local star-forminggalaxies and active galactic nuclei (AGN) were first car- † Contact email: [email protected] ried out with
ISO (e.g., Malhotra et al. 1997, 2001; Luh-man et al. 1998; Brauher et al. 2008). These studiesshowed that [C ii ]157.7 µ m is the most intense far-IRemission line observed in normal, star-forming galax-ies (Malhotra et al. 1997) and starbursts (e.g., Nikolaet al. 1998; Colbert et al. 1999), dominating the gascooling of their neutral inter stellar medium (ISM). D´ıaz-Santos et al.This fine-structure line arises from the P / → P / transition ( E ul /k = 92 K) of singley ionized Carbonatoms (ionization potential = 11.26 eV and critical den-sity, n crH ≃ . × cm − ; n cre − ≃
46 cm − ) which arepredominantly excited by collisions with neutral hydro-gen atoms; or with free electrons and protons in regionswhere n e − /n H & − (Hayes & Nussbaumer 1984). Ul-traviolet (UV) photons with energies > + atoms and other elementsin photo-dissociation regions (PDRs) (Tielens & Hollen-bach 1985; Wolfire et al. 1995).The [C ii ]157.7 µ m emission accounts, in the most ex-treme cases, for as much as ∼ ii ]/FIR ratio is observed to decrease bymore than an order of magnitude in sources with high L IR and warm dust temperatures ( T dust ). The underly-ing causes for these trends are still debated. The physicalarguments most often proposed to explain the decreasein [C ii ]/FIR are: (1) self-absorption of the C + emission,(2) saturation of the [C ii ] line flux due to high density ofthe neutral gas, (3) progressive ionization of dust grainsin high far-UV field to gas density environments, and(4) high dust-to-gas opacity caused by an increase of theaverage ionization parameter.Although self absorption has been used to explain thefaint [C ii ] emission arising from warm, AGN-dominatedsystems such as Mrk 231 (Fischer et al. 2010), this in-terpretation has been questioned in normal star-forminggalaxies due to the requirement of extraordinarly largecolumn densities of gas in the PDRs (Luhman et al. 1998,Malhotra et al. 2001). Furthermore, contrary to the [O i ]or [C i ] lines, the [C ii ] emission is observed to arise fromthe external edges of those molecular clouds exposed tothe UV radiation originated from starbursts, as for exam-ple in Arp 220 (Contini 2012). Therefore, self absorptionis not the likely explanation of the low [C ii ]/FIR ratiosseen in most starburst galaxies, except perhaps in a fewextreme cases, like NGC 4418 (Malhotra et al. 1997).The [C ii ] emission becomes saturated when the hydro-gen density in the neutral medium, n H , increases to val-ues & cm − , provided that the far-UV (6 − . G . ; where G is nor-malized to the average local interstellar radiation field;Habing 1968). For example, for a constant G = 10 ,an increase of the gas density from 10 to 10 cm − would produce a suppression of the [C ii ] emission ofalmost 2 orders of magnitude due to the rapid recom-bination of C + into neutral Carbon and then into CO(Kaufman et al. 1999). However, PDR densities as highas 10 cm − are not very common. [O i ]63.18 µ m and[C ii ]157.7 µ m ISO observations of normal star-forminggalaxies and some IR-bright sources confine the physi-cal parameters of their PDRs to a range of G ≤ . and 10 . n H . cm − (Malhotra et al. 2001). Onthe other hand, the [C ii ] emission can be also saturated when G > . provided that n H . cm − . In thisregime, the line is not sensitive to an increase of G be-cause the temperature of the gas is well above the exci-tation potential of the [C ii ] transition.It has also been suggested that in sources where G / n H is high ( & cm ) the [C ii ] line is a less efficient coolantof the ISM because of the following reason. As physicalconditions become more extreme (higher G / n H ), dustparticles progressively increase their positive charge (Tie-lens & Hollenbach 1985; Malhotra et al. 1997; Negishiet al. 2001). This reduces both the amount of photo-electrons released from dust grains that indirectly colli-sionally excite the gas, as well as the energy that theycarry along after they are freed, since they are morestrongly bounded. The net effect is the decreasing ofthe efficiency in the transformation of incident UV radi-ation into gas heating without an accompanied reductionof the dust emission (Wolfire et al. 1990; Kaufman et al.1999; Stacey et al. 2010).In a recent work, Graci´a-Carpio et al. (2011) haveshown that the deficits observed in several far-IR emis-sion lines ([C ii ]157.7 µ m, [O i ]63.18 µ m, [O i ]145 µ m, and[N ii ]122 µ m) could be explained by an increase of the av-erage ionization parameter of the ISM, < U > ∗ . In ”dustbounded” star-forming regions the gas opacity is reducedwithin the H ii region due to the higher < U > . As aconsequence, a significant fraction of the UV radiationis eventually absorbed by large dust grains before beingable to reach the neutral gas in the PDRs and ionize thePAH molecules (Voit 1992; Gonz´alez-Alfonso et al. 2004;Abel et al. 2009), causing a deficit of photo-electrons andhence the subsequent suppression of the [C ii ] line withrespect to the total far-IR dust emission.Local luminous infrared galaxies (LIRGs: L IR =10 − L ⊙ ) are a mixture of single galaxies, disk galaxypairs, interacting systems and advanced mergers, ex-hibiting enhanced star formation rates, and a lower frac-tion of AGN compared to higher luminous galaxies. Adetailed study of the physical properties of low-redshiftLIRGs is critical for our understanding of the cosmic evo-lution of galaxies and black holes since (1) IR-luminousgalaxies comprise the bulk of the cosmic infrared back-ground and dominate star-formation activity between0 . < z < z >
1. However, a comprehensive analysisof the most important far-IR cooling lines of the ISM ina complete sample of nearby LIRGs has not been possi-ble until the advent of the
Herschel Space Observatory ( Herschel hereafter; Pilbratt et al. 2010) and, in partic-ular, its Photodetector Array Camera and Spectrometer(PACS; Poglitsch et al. 2010).In this work we present the first results obtained from
Herschel /PACS spectroscopic observations of a completesample of far-IR selected local LIRGs that comprise the ∗ The ionization parameter is defined as U ≡ Q (H) / πR n H c ,where Q (H) is the number of hydrogen ionizing photons, R is thedistance of the ionizing source to the PDR, n H is the atomic hy-drogen density, and c is the speed of light. If an average stellarpopulation and size for the star-forming region is assumed, then U ∝ G / n H . xplaining the [C ii ] Deficit in LIRGs – First GOALS Results From Herschel/PACS 3Great Observatories All-sky LIRG Survey (GOALS; Ar-mus et al. 2009). Using this complete, flux-limited sam-ple of local LIRGs, we are able for the first time to per-form a systematic, statistically significant study of thefar-IR cooling lines of star-forming galaxies covering awide range of physical conditions: from isolated diskswhere star formation is spread across kpc scales to themost extreme environments present in late stage majormergers where most of the energy output of the systemcomes from its central kpc region. In particular, in thispaper we focus on the [C ii ]157.7 µ m line and its rela-tion with the dust emission in LIRGs. We make use ofa broad set of mid-IR diagnostics based on Spitzer /IRSspectroscopy, such as high ionization emission lines, sil-icate dust opacities, PAH equivalent widths (EW), dustluminosity concentrations, and mid-IR colors, to providethe context in which the observed [C ii ] emission and[C ii ]/FIR ratios are best explained. The paper is or-ganized as follows: In § § § § § SAMPLE AND OBSERVATIONS
The GOALS Sample
The Great Observatories All-sky LIRG Survey(GOALS; Armus et al. 2009) encompasses the completesample of 202 LIRGs and ULIRGs contained in the
IRAS
Revised Bright Galaxy Sample (RBGS; Sanderset al. 2003) which, in turn, is also a complete sam-ple of 629 galaxies with
IRAS S µm > .
24 Jy andGalactic latitudes | b | > ◦ . There are 180 LIRGs and22 ULIRGs in GOALS and their median redshift is z =0.0215 (or ∼ . HST optical and near-IR (Haanet al. 2011; Kim et al. 2013), and
Chandra
X-ray (Iwa-sawa et al. 2011) imaging, as well as
Spitzer /IRS mid-IR spectroscopy (D´ıaz-Santos et al. 2010b, 2011; Petricet al. 2011; Stierwalt et al. 2013a,b; Inami et al. 2013),as well as a number of ground-based observatories (VLA,CARMA, etc.) and soon ALMA.The RBGS, and therefore the GOALS sample, were de-fined based on
IRAS observations. However, the higherangular resolution achieved by
Spitzer allowed us to spa-tially disentangle galaxies that belong to the same LIRGsystem into separate components. From the 291 individ-ual galaxies in GOALS, not all have
Herschel observa-tions. In systems with two or more galactic nuclei, minorcompanions with MIPS 24 µ m flux density ratios smallerthan 1:5 with respect to the brightest galaxy were notrequested since their contribution to the total IR lumi-nosity of the system is small. Because the angular res-olution of Spitzer decreases with wavelength, it was notpossible to obtain individual MIPS 24, 70 and 160 µ mmeasurements for all GOALS galaxies, and therefore to derive uniform IR luminosities for them using Spitzer data only. Instead, to calculate the individual, spatially-integrated L IR of LIRGs belonging to a system of two ormore galaxies, we distributed the L − µ mIR of the sys-tem as measured by IRAS (using the prescription givenin Sanders & Mirabel 1996) proportionally to the indi-vidual MIPS 70 µ m flux density of each component whenavailable, or to their MIPS 24 µ m otherwise . We willuse this measurements of L IR in § Herschel/PACS Observations
We have obtained far-IR spectroscopic observations for153 LIRG systems of the GOALS sample using the In-tegral Field Spectrometer (IFS) of the PACS instrumenton board
Herschel . The data were collected as part of anOT1 program (OT1 larmus 1; P.I.: L. Armus) awardedwith more than 165 hours of observing time. In thiswork will focus mainly on the analysis and interpretationof the [C ii ] observations of our galaxy sample. PACSrange spectroscopy of the [C ii ]157.7 µ m fine-structureemission line was obtained for 163 individual sources.Our observations were complemented with the inclusionof the remaining LIRGs in the GOALS sample for which[C ii ] observations are publicly available in the archive(as of October 2012) from various Herschel projects. Themain programs from which these data were gathered are:KPGT esturm 1 (P.I.: E. Sturm), KPOT pvanderw 1(P.I.: P. van der Werf), and OT1 dweedman 1 (P.I.:D. Weedman). The total number of LIRG systems forwhich there are [C ii ] data is 200 (IRASF08339+6517and IRASF09111-1007 were not observed). However, be-cause some LIRGs are actually systems of galaxies (seeabove), the number of observed galaxies was 241.The IFS on PACS is able to perform simultaneous spec-troscopy in the 51 −
73 or 70 − µ m (3rd and 2nd or-ders, respectively; ”blue” camera) and the 102 − µ m(1st order; ”red” camera) ranges. The IFU is composedby a 5 × ∼ ′′ , for a to-tal of 47 ′′ × ′′ . The physical size of the PACS FoV atthe median distance of our LIRG sample is ∼
20 kpc ona side. The number of spectral elements in each pixelis 16, which are rearranged together via an image slicerover two 16 ×
25 Ge:Ga detector arrays (blue and redcameras).Our Astronomical Observation Requests (AORs) wereconsistently constructed using the ”Range” spectroscopytemplate, which allows the user to define a specific wave-length range for the desired observations. Our selectedrange was slightly larger than that provided by defaultfor the ”Line” mode. This was necessary (1) to obtainparallel observations of the wide OH 79.18 µ m absorp-tion feature using the blue camera when observing the[C ii ]157.7 µ m line, and (2) to ensure that the targetedemission lines have a uniform signal-to-noise ratio acrosstheir spectral profiles even if they are to be broader thana few hundred km s − . The high sampling density modescan, useful to have sub-spectral resolution informationof the lines (see below), was employed. While we re- There are two systems for which no individual MIPS 24 µ mfluxes could be obtained. In these cases their IRAC 8 µ m emissionwas used for scaling the L IR . These LIRGs are: MCG+02-20-003and VV250a. D´ıaz-Santos et al.quested line maps for some LIRGs of the sample (fromtwo to a few raster positions depending on the target),pointed (one single raster) chop-nod observations weretaken for the majority of galaxies. For those galaxieswith maps, only one raster position was used to obtainthe line fluxes used in this work. The chopper throw var-ied from small to large depending on the source. Spec-troscopy of the LIRGs included in GOALS but observedby other programs in [C ii ] was not always obtained us-ing the ”Range” mode but some of them were observedusing ”LineScan” spectroscopy. The S/N of the datavaries not only from galaxy to galaxy but also dependingon the emission line considered. We provide uncertain-ties for all quantities used across the analysis presentedhere that are based on the individual spectrum of eachline, therefore reflecting the errors associated with − andmeasured directly on − the data. Spitzer/IRS Spectroscopy
As part of the
Spitzer
GOALS legacy, all galaxies ob-served with
Herschel /PACS have available
Spitzer /IRSlow resolution (R ∼ − . − . µ m, and LL module: 14 − µ m). The244 IRS spectra were extracted using the standard ex-traction aperture and point source calibration mode inSPICE. The projected angular sizes of the apertures onthe sky are 3.7 ′′ × ′′ at the average wavelength of 10 µ min SL and 10.6 ′′ × ′′ at the average wavelength of 26 µ min LL. Thus, the area covered by the SL aperture is ap-proximately equivalent (within a factor of ∼
2) to that ofan individual spaxel of the IFS in PACS, and so is thatof the LL aperture to a 3 × µ m silicate feature, S . µ m ,and the EW of the 6.2 µ m PAH, which were presentedin Stierwalt et al. (2013a). We refer the reader to thiswork for further details about the reduction, extraction,calibration, and analysis of the spectra. IFS/PACS DATA REDUCTION AND ANALYSIS
Data Processing
The
Herschel
Interactive Processing Environment(HIPE; v8.0) application was used to retrieve the rawdata from the
Herschel
Science Archive (HSA ) as well asto process them. We used the script for ”LineScan” ob-servations (also valid for ”Range” mode) included withinHIPE to reduce our spectra. We processed the data fromlevel 0 up to level 2 using the following steps: Flag andreject saturated data, perform initial calibrations, flagand reject ”glitches”, compute the differential signal ofeach on-off pair of data-points for each chopper cycle,calculate the relative spectral response function, divideby the response, convert frames to PACS cubes, and cor-rect for flat-fielding (this extra step is included in v8.0of HIPE and later versions, and helps to improve theaccuracy of the continuum level). Next, for each cam-era (red or blue), HIPE builds the wavelength grid, forwhich we chose a final rebinning with an oversample = 2,and an upsample = 3 that corresponds to a Nyquist sam-pling. The spectral resolution achieved at the positionof the [C ii ]157.7 µ m line was derived directly from the http://herschel.esac.esa.int/Science Archive.shtml data and is ∼
235 km s − . The final steps are: flag andreject remaining outliers, rebin all selected cubes on con-sistent wavelength grids and, finally, average the nod-Aand nod-B rebinned cubes (all cubes at the same rasterposition are averaged). This is the final science-gradeproduct currently possible for single raster observations.From this point on, the analysis of the spectra was per-formed using in-house developed IDL routines. Data Analysis
To obtain the [C ii ] flux of a particular source we usean iterative procedure to find the line and measure itsbasic parameters. First, we fit a linear function to thecontinuum emission, which is evaluated at the edges ofthe spectrum, masking the central 60 % of spectral ele-ments (where the line is expected to be detected) andwithout using the first and final 10 %, where the noise islarge due to the poor sampling of the scanning. Then,we fit a Gaussian function to the continuum-subtractedspectrum and calculate its parameters. We define a lineas not detected when the peak of the Gaussian is below2.5 × the standard deviation of the continuum, as mea-sured in the previous step. On the other hand, if the lineis found, we return to the original, total spectrum and fitagain the continuum using this time a wavelength rangedetermined by the two portions of the spectrum adjacentto the line located beyond ± σ from its center (where σ is the width of the fitted Gaussian) and the following ±
15 % of spectral elements. We then subtract this con-tinuum from the total spectrum and fit the line again.The new parameters of the Gaussian are compared withthe previous ones. This process is repeated until the loca-tion, sigma and intensity of the line converge with an ac-curacy of 1 %, or when reaching 10 iterations. Due to themerger-driven nature of many LIRGs, their gas kinemat-ics are extremely complicated and, as a consequence, theemission lines of several sources present asymmetries anddouble peaks in their profiles. However, despite the factthat the width determined by the fit is not an accuraterepresentation of the real shape of the line, it can be usedas a first order approximation for its broadness. There-fore, instead of using the parameters of the Gaussian toderive the flux of the line, we decided to integrate directlyover the final continuum-subtracted spectrum within the ± σ region around the central position of the line. Theassociated uncertainty is calculated as the standard devi-ation of the latest fitted continuum, integrated over thesame wavelength range as the line. Absolute photomet-ric uncertainties due to changes in the PACS calibrationproducts are not taken into account (the version used inthis work was PACS CAL 32 0) .We obtained the line fluxes for our LIRGs from thespectra extracted from the spaxel at which the [C ii ] line+ continuum emission of each galaxy peaks within thePACS FoV. The Spitzer /IRS and
Herschel pointings usu-ally coincide within . ′′ . There are a few targets forwhich the IRS pointing is located more than half a spaxelaway from that of PACS. In these cases, we decided toobtain the nuclear line flux of the galaxy by averagingthe spaxels closest to the coordinates of the IRS point-ing. These values are used only when PACS and IRSmeasurements are compared directly in the same plot. http://herschel.esac.esa.int/twiki/bin/view/Public/PacsCalTreeHistory xplaining the [C ii ] Deficit in LIRGs – First GOALS Results From Herschel/PACS 5There is one additional LIRG system, IRAS03582+6012,for which the PACS pointing exactly felt in the middle oftwo galaxies separated by only 5 ′′ . This LIRG is not usedin the comparisons of the [C ii ] emission to the IRS datasince the two individual sources cannot be disentangled.As mentioned in § Spizer /IRS spectra of ourgalaxies. Because the PACS beam is under-sampled at160 µ m (FWHM ∼ ′′ compared with the 9 . ′′ size ofthe PACS spaxels), and most of the sources in the sam-ple are unresolved at 24 µ m in our MIPS images (whichhave a similar angular resolution as PACS at ∼ µ m),an aperture correction has to be performed to the spectraextracted from the emission-peak spaxel of each galaxyto obtain their total nuclear fluxes. This was the sameprocedure employed to obtain the mid-IR IRS spectra ofour LIRGs. The nominal, wavelength-dependent aper-ture correction function provided by HIPE v8.0 worksoptimally when the source is exactly positioned at thecenter of a given spaxel. However, in some occasions thepointing of Herschel is not accurate enough to achievethis and the target can be slightly misplaced . ′′ (upto 1/3 of a spaxel) from the center. In these cases, theflux of the line might be underestimated. We exploredwhether this effect could be corrected by measuring theposition of the source within the spaxel. However, someLIRGs in our sample show low surface-brightness ex-tended emission, either because of their proximity and/ormerger nature, or simply because the gas and dust emis-sion are spatially decoupled. This, combined with thespatial sub-sampling of the PACS/IFS detector and thepoor S/N of some sources prevented us from obtaining anaccurate measurement of the spatial position and angularwidth of the [C ii ] emission and therefore from obtaininga more refined aperture correction. Thus, we performedonly the nominal aperture correction provided by HIPE.The IRAS far-IR fluxes used throughout this paperwere calculated as FIR = 1 . × − (2 . S µ m + S µ m ) [W m − ], with S ν in [Jy]. The far-IR luminosi-ties, L FIR , were defined as 4 π D FIR [ L ⊙ ]. The lu-minosity distances, D L , were taken from Armus et al.(2009). This definition of the FIR accounts for the fluxemitted within the 42 . − . µ m wavelength range asoriginally defined in Helou et al. (1988). The far-IR fluxesand luminosities of galaxies were then matched to theaperture with which the nuclear [C ii ] flux was extracted(see above) by scaling the integrated IRAS far-IR fluxof the LIRG system with the ratio of the continuum fluxdensity of each individual galaxy evaluated at 63 µ m inthe PACS spectrum (extracted at the same position andwith the same aperture as the [C ii ] line) to the total IRAS µ m flux density of the system.In Table 1 we present the [C ii ]157.7 µ m flux, the[C ii ]/FIR ratio, and the continuum flux densities at 63and 158 µ m for all the galaxies in our sample. Future up-dates of the data in this table processed with newer ver-sions of HIPE and PACS calibration files will be availableat the GOALS webpage: http://goals.ipac.caltech.edu. RESULTS AND DISCUSION
The [C ii ]/FIR Ratio: Dust Heating and Cooling Fig. 1.—
The ratio of [C ii ]157.7 µ m to far-IR flux as a function ofthe far-IR luminosity for individual galaxies in the GOALS sample(green circles) and for unresolved galaxies observed with ISO (graycircles and limits) obtained from the compilation of Brauher et al.(2008) located at z > . L FIR of galaxies was calculated as explainedat the end of § . − . µ m wavelength rangeas defined in Helou et al. (1988). The far-IR fine structure line emission in normal star-forming galaxies as well as in the extreme environmentshosted by ULIRGs has been extensively studied for thepast two decades. A number of works based on
ISO data already suggested that the relative contribution ofthe [C ii ]157.7 µ m line to the cooling of the ISM in PDRscompared to that of large dust grains, as gauged by thefar-IR emission, diminishes as galaxies are more IR lumi-nous (Malhotra et al. 1997; Luhman et al. 1998; Brauheret al. 2008). Figure 1 display the classical plot of the[C ii ]/FIR ratio as a function of the far-IR luminosityfor our LIRG sample. In addition, we also show forreference those galaxies observed with ISO compiled byBrauher et al. (2008) that are classified as unresolved andlocated at redshifts z > . Herschel data confirm the trend seen with
ISO by which galaxieswith L FIR & L ⊙ show a significant decrease of the[C ii ]/FIR ratio. GOALS densely populates this criticalpart of phase-space providing a large sample of galaxieswith which to explore the physical conditions behind thedrop in [C ii ] emission among LIRGs. For the 32 galaxieswith measurements obtained with both telescopes, thehigher angular resolution Herschel observations of thenuclei of LIRGs are able to recover an average of ∼ ii ] flux measured by ISO . The Average Dust Temperature of LIRGs
Figure 2 (upper panel) shows the [C ii ]157.7 µ m/FIRratio for the GOALS sample as a function of the far-IRPACS S ν µ m/ S ν µ m continuum flux density ratio.We chose to use this PACS-based far-IR color in the x-axis instead of the more common IRAS µ m colormainly because of two main reasons: (1) this way weare able plot data from individual galaxies instead of be-ing constrained by the spatial resolution of IRAS , whichwould force us to show only blended sources; (2) by using D´ıaz-Santos et al.
TABLE 1
Herschel /PACS measurements for the GOALS sample
Galaxy R.A. Dec. Dist. [C ii ]157.7 µ m [C ii ]/FIR S ν µ m cont. S ν µ m cont.name [hh:mm:ss] [dd:mm:ss] [Mpc] [ × − W m − ] [ × − ] [Jy] [Jy](1) (2) (3) (4) (5) (6) (7) (8)NGC0023 00h09m53.4s +25 ◦ ± ± ± ± − ◦ ± ± ± ± − ◦ ± ± ± ± Note . — (1) Galaxy name; (2) − (3) Right Ascension and Declination (J2000) of the position from which the Herschel /PACS spectrum wasextracted ( § ii ]157.7 µ m flux as measured from the spaxel at which the [C ii ]line + continuum emission of the galaxy peaks within the PACS FoV; that is, within an effective aperture of ∼ . ′′ × . ′′ ( § ii ] tofar-IR flux ratio, where the far-IR fluxes have been scaled to match the aperture of the [C ii ] measurements; (7) Continuum flux density at 63 µ munder the [O i ] line extracted at the same position and with the same aperture size as (5); (8) Same as (7) but for the continuum at 158 µ m underthe [C ii ] line.There are 11 galaxies for which the Spitzer /IRS pointing is located > . ′′ from the position of the [C ii ]+continuum peak. For these, we includean extra entry in the table with the [C ii ] flux and continuum measurements obtained at the position of the IRS slit. They are marked with asterisksnext to the names.The complete table is available in the electronic edition of the paper. Future updates of the data in this table processed with newer versions ofHIPE and PACS calibration les will be available at the GOALS webpage: http://goals.ipac.caltech.edu the 63/158 µ m ratio we are probing a larger range of dusttemperatures within the starburst ( T ∼ →
20 K);with the colder component probably arising from regionslocated far from the ionized gas-phase, and closer to thePDRs where the [C ii ] emission originates. For reference,we show the relation between the PACS 63/158 µ m and IRAS µ m colors in the Appendix. The ULIRGsin the GOALS sample (red diamonds) have a median[C ii ]/FIR = 6 . × − , a mean of 6 . ± . × − , anda standard deviation of the distribution of 3 . × − .LIRGs span two orders of magnitude in [C ii ]/FIR, from ∼ − to ∼ − , with a mean of 3 . × − and amedian of 2 . × − . The L [C II] ranges from ∼ to2 × L ⊙ .Our results are consistent with ISO observations ofa sample of normal and moderate IR-luminous galax-ies presented in Malhotra et al. (1997) and further an-alyzed in Helou et al. (2001). The GOALS sample,though, populates a warmer far-IR color regime. De-spite the increase in dispersion at 63/158 µ m & .
25 or[C ii ]157.7 µ m/FIR . − (basically in the ULIRG do-main), the fact that we find the same tight trend inde-pendently of the range of IR luminosities covered by thetwo samples suggests that the main observable linked tothe variation of the [C ii ]/FIR ratio is the average tem-perature of the dust ( T dust ) in galaxies.This interpretation agrees with the last physical sce-nario described in the Introduction, in which an increaseof the ionization parameter, < U > , would cause the far-UV radiation from the youngest stars to be less efficientin heating the gas in those galaxies. At the same time,dust grains would be on average at higher temperaturesdue to the larger number of ionizing photons per dustparticle available in the outer layers of the H ii regions,close to the PDRs. Indeed, the presence of dust withinH ii regions has been recently observed in several star-forming regions in our Galaxy (Paladini et al. 2012).Both effects combined can explain the wide range of[C ii ]157.7 µ m/FIR ratios and far-IR colors we observe inthe most warm LIRGs. Variations in n H , though, couldbe responsible for the dispersion in [C ii ]/FIR seen at agiven 63/158 µ m ratio.To further support these findings, the bottom panel of Figure 2 shows that the ratio of [C ii ]157.7 µ m fluxto the monochromatic continuum at ∼ µ m under theline (the [C ii ] EW) of the warmest galaxies is only afactor of ∼ ii ] EW displayedby colder sources at 63/158 µ m .
1. This implies thatthe decrease of the [C ii ]/FIR ratio seen in our LIRGs isprimarily caused by a significant increase in warm dustemission (peaking at λ . − µ m), most likely as-sociated with the youngest stars, that is not followed bya proportional enhancement of the [C ii ] emission line.The best fit to the data in Figure 2 (upper panel) yieldsthe following parameters: log ( [C ii ]FIR ) = − . ± . − . ± . log ( S µm S µm )(1)with a dispersion of 0.28 dex. We note that the [C ii ]/FIRratios predicted by the fitted relation for sources with far-IR colors 63/158 µ m . . T dust have typical [C ii ]/FIR ∼ − (e.g., Malhotra et al.2001). The Link Between Mid-IR Dust Obscuration andFar-IR Re-emission
The strength of the 9.7 µ m silicate feature is defined as S . µ m ≡ ln ( f obsλ P /f contλ P ); with f contλ P and f obsλ P being theun-obscured and observed continuum flux density mea-sured in the mid-IR IRS spectra of our LIRGs and eval-uated at the peak of the feature, λ P , normally at 9.7 µ m(see Stierwalt et al. 2013a for details on how it was cal-culated in our sample). Negative values indicate absorp-tion, while positive ones indicate emission. By definition, S . µ m measures the apparent optical depth towards thewarm, mid-IR emitting dust. Figure 3 shows that thereis a clear trend ( r = 0 . , p r = 0; κ = 0 .
47) for LIRGswith stronger (more negative) S . µ m to display smaller[C ii ]/FIR ratios, implying that the dust responsible forthe mid-IR absorption is also accountable for the far-IRemission. The formal fit (solid line) can be expressed as: log ( [C ii ]FIR ) = − . ± .
02) + 0 . ± . S . µ m (2)xplaining the [C ii ] Deficit in LIRGs – First GOALS Results From Herschel/PACS 7 Fig. 2.—
The ratio of [C ii ]157.7 µ m to far-IR flux (upperpanel) and [C ii ]157.7 µ m EW (bottom panel) as a function ofthe S ν µ m/ S ν µ m continuum flux density ratio for individualgalaxies in the GOALS sample. Circles of different colors indicatethe L FIR of galaxies (see color bar), which is defined as explained in § L IR ≥ L ⊙ , ULIRGs.These two plots show that the decrease of the [C ii ]/FIR ratio withwarmer far-IR colors seen in our LIRGs is primarily caused by a sig-nificant decrease of gas heating efficiency and an increase of warmdust emission. The solid line in the upper panel corresponds to alinear fit of the data in log-log space. The parameters of the fit aregiven in Eq. (1). The dotted lines are the ± σ uncertainty. with a dispersion in the y-axis of 0.30 dex.Within the context described in the previous section,the contrast between the inner layer of dust that is beingheated by the ionizing radiation to T &
50 K and thatof the cold dust at T .
20 K emitting at λ & µ mwould create both: (1) the silicate absorption seen at9.7 µ m due to the larger temperature gradient betweenthe two dust components and (2) the increasingly higher63/158 µ m ratios seen in Figure 2 due to the progres-sively larger amount of dust mass that is being heated tohigher temperatures. This scenario is consistent with thephysical properties of the ISM found in the extreme envi-ronments of ULIRGs, in which the fraction of total dustluminosity contributed by the diffuse ISM decreases sig-nificantly, and the emission from dust at T ∼ −
60 K
Fig. 3.— [C ii ]157.7 µ m/FIR ratio for individual galaxies in theGOALS sample as a function of the strength of the 9.7 µ m silicateabsorption feature, S . µ m , measured with Spitzer /IRS. Galaxiesare color-coded as a function of the S ν µ m/ S ν µ m ratio, withvalues ranging from 0.10 to 4.15 (see Figure 2). The solid linerepresents an outlier-resistant fit to the bulk of the star-formingLIRG population with S . µ m ≥ −
2. The dotted lines are the ± σ uncertainty. arising from optically-thick ”birth clouds” (with ages . − Myr) accounts for &
80% of their IR energy out-put (da Cunha et al. 2010). Furthermore, our findingsare also in agreement with recent results showing thatthe increase of the silicate optical depth in LIRGs is re-lated with the flattening of their radio spectral index(1.4 to 8.44 GHz) due to an increase of free-free absorp-tion, suggesting that the dust obscuration must largelybe originated in the vicinity and/or within the starburstregion (Murphy et al. 2013).There are a few galaxies that do not follow the cor-relation fitted in Figure 3, showing very large silicatestrengths ( S . µ m < −
2) and small [C ii ]/FIR ratios( < − ) typical of ULIRGs or, in general, warm galax-ies (see color coding or Figure 2). We would like to notethat the trend found for the majority of our sample onlyreaches S . µ m values up to around − . ii ] and far-IR emissions arise.But then, what is the origin of this excess of obscu-ration? One possibility is that it is caused by fore-ground cold dust not associated with the starburst. Thishas been seen in some heavily obscured Compton-thickAGNs, where most of the deep silicate absorption mea-sured in these objects seems to originate from dust lo-cated in the host galaxy (Goulding et al. 2012; Gonzalez-Martin et al. 2012). Alternatively, the presence of anextremely warm source (different from the star-formingregion(s) that are producing the far-IR and [C ii ] emis-sion) could contribute with additional emission of hotdust (T &
150 K) to the mid-IR. If at the same time this D´ıaz-Santos et al.source is deeply buried (optically thick) and embedded inlayers of progressively colder dust (geometrically thick),it could produce a cumulative absorption that we wouldmeasure via the strength of the silicate feature while stillcontributing to the emission outside of it (see Levensonet al. 2007; Sirocky et al. 2008).While both explanations are plausible, the second is fa-vored by the fact that these extremely obscured galaxiesshow MIPS 24/70 µ m ratios very similar, or even slightlyhigher than those found for the rest of the LIRGs inthe sample. If foreground cold dust was the respon-sible for the excess of obscuration, we would expectthese galaxies to show abnormally low 24 µ m luminosi-ties with respect to the far-IR. We find that this is notthe case, in agreement with recent results based on ra-dio observations of a sub-sample of LIRGs in GOALS(Murphy et al. 2013). Furthermore, the existence ofan additional hot and obscured dust component in theseLIRGs is also consistent with the results presented inStierwalt et al. (2013a), where it is shown that thereis a trend for LIRGs with moderate silicate strengths( S . µ m & − .
5) to show higher S ν µ m/ S ν µ m ra-tios as the S . µ m becomes stronger (more negative).That is, more obscured LIRGs have increasingly largerfluxes at 30 µ m, in agreement with our findings in theprevious section. However, galaxies showing the mostextreme silicate strengths ( S . µ m < − .
5) do not haveproportionally higher 30/15 µ m ratios. On the contrary,they show ratios similar to those of warm LIRGs withmild silicate strengths (or even lower than expected giventheir extreme S . µ m ), supporting the idea that in theseparticular galaxies the dust producing this additional ab-sorption and excess of mid-IR emission ( λ . µ m)represents a component of the overall nuclear starburstactivity different than the star-forming regions that drivethe far-IR cooling. The Compactness of the Mid-IR Emitting Region
The compactness of the starburst region of a galaxyhas been proven to be related to many of its otherphysical properties (Wang & Helou 1992). For exam-ple, all ULIRGs in the GOALS sample have very smallmid-IR emitting regions, with sizes (measured FWHMs) < . λ , which mea-sures the fraction of light emitted by a galaxy that iscontained outside of its unresolved component at a givenwavelength λ . The complementary quantity 1 − FEE λ measures how compact the source is, which in turn isproportional to its luminosity surface density, Σ. Wenote that in this paper we use the word compactness asan equivalent to light concentration, i.e., as a measure-ment of the amount of energy per unit area produced bya source, and not as an absolute measurement of its size.It has been shown that the compactness of the mid-IR continuum emission of LIRGs (evaluated at λ rest =13 . µ m) is related to their merger stage, mid-IR AGN-fraction and most importantly, to their far-IR color(D´ıaz-Santos et al. 2010b). LIRGs with higher IRAS S ν µ m/ S ν µ m ratios are increasingly compact. Inother words, for a given L IR , the dust in sources with Fig. 4.— [C ii ]157.7 µ m/FIR ratio as a function of the lumi-nosity surface density at 15 µ m, Σ µ m (top), and the fraction ofextended emission at 13.2 µ m, FEE . µm (bottom), for individ-ual galaxies in the GOALS sample. Galaxies are color-coded asa function of the 63/158 µ m ratio (top) and the strength of thesilicate feature, S . µ m (bottom). A linear fit to the datapointswithout limits is shown in the top panel as a solid line. The dottedlines are the ± σ uncertainty. We note that the 15 µ luminosi-ties are measured within the Spitzer /IRS LL slit while the mid-IRsizes were obtained from the SL module at 13.2 µ m (D´ıaz-Santoset al. 2010b). For very extended sources, the MIPS 24 µ m imageswere used instead to measure the size of the starburst region. Theintrinsic sizes (FWHM int ) of the mid-IR emission were obtained af-ter subtracting, in quadrature, the contribution of the instrumentalprofile (FWHM PSF ) from the measured FWHM. far-IR colors peaking at shorter wavelengths is not onlyhotter but also confined towards a smaller volume in thecenter of galaxies. In § ii ]/FIRratio is related to the average T dust of our galaxies. Thus,we should expect to see a correlation between the [C ii ]deficit and the luminosity surface density and compact-ness of LIRGs in the mid-IR. This is shown in Figure 4,where a clear trend is found for galaxies with higher lu-minosity surface densities at 15 µ m, Σ µ m (top panel),or small FEE . µm (bottom panel), i.e, more compact,to show lower [C ii ]/FIR ratios, irrespective of the originof the nuclear power source.Excluding those LIRGs for which only upper limits ontheir mid-IR size or L [C II] are available, we perform axplaining the [C ii ] Deficit in LIRGs – First GOALS Results From Herschel/PACS 9linear fit to the data and obtain the following parametersfor the correlation between [C ii ]/FIR and Σ µ m : log ( [C ii ]FIR ) = 0 . ± . − . ± . × log (Σ µ m )(3)with a dispersion in the y-axis of 0.23 dex. The Role of Active Galactic Nuclei
It is known that the contribution of an AGN to theIR emission in LIRGs increases with L IR (Veilleux et al.1995; Desai et al. 2007; Petric et al. 2011; Alonso-Herreroet al. 2012). This is most noticeable at mid-IR wave-lengths (Laurent et al. 2000; Armus et al. 2007; Mullaneyet al. 2011) but a non-negligible fraction of the far-IRemission of ULIRGs can also be powered by an AGN.The EW of mid-IR PAH features is a simple diagnosticthat has been widely used for the detection of AGN ac-tivity in galaxies at low and high redshifts (Genzel et al.1998; Armus et al. 2007; Desai et al. 2007; Spoon et al.2007; Pope et al. 2008; Murphy et al. 2009; Men´endez-Delmestre et al. 2009; Veilleux et al. 2009; Petric et al.2011; Stierwalt et al. 2013a). Broadly speaking, the PAHEW decreases as a component of hot dust at T &
300 K,normally ascribed to an AGN, starts to increasingly dom-inate the mid-IR continuum emission of the galaxy. Inaddition, the hard radiation field of an AGN could beable to destroy a significant fraction of the smallest PAHmolecules (Voit 1992; Siebenmorgen et al. 2004). In par-ticular, a galaxy is regarded as mid-IR AGN-dominatedwhen its 6.2 µ m PAH EW . .
3, and it is classified asa pure starburst when 6.2 µ m PAH EW & . [C ii ]157.7 µ m Deficit in Pure Star-Forming LIRGs Figure 5 shows the [C ii ]/FIR ratio as a function ofthe IR luminosity surface density, Σ IR , of the LIRGs inGOALS, color-coded as a function of their 6.2 µ m PAHEW. As we can see, when only pure star-forming galax-ies are considered (6.2 µ m PAH EWs ≥ . µ m), the[C ii ]/FIR ratio drops by an order of magnitude, from10 − to ∼ − . This indicates that the decrease in[C ii ]/FIR among the majority of LIRGs is not causedby a rise of AGN activity but instead is a fundamentalproperty of the starburst itself. It is only in the mostextreme cases, when [C ii ]/FIR < − , that the AGNcould play a significant role. In fact, powerful AGN donot always reduce the [C ii ]/FIR ratio, as shown also inSargsyan et al. (2012). Stacey et al. (2010) also find thatthe AGN-powered sources in their high-redshift galaxysample display small [C ii ]/FIR ratios. However, theyspeculate that except for two blazars, the deficit seenin these sources could be caused compact, nuclear star-bursts (with sizes less than 1 − § ii ]157.7 µ m line alone is not a good tracer of the SFRin most local LIRGs since it does not account for the in-crease of warm dust emission (Figure 2) seen in the most Fig. 5.— [C ii ]157.7 µ m/FIR ratio as a function of the nu-clear L IR divided by the area of the mid-IR emitting region(Σ IR = L IR /πr − IR ) for individual galaxies in the GOALSsample. This Figure is the same as Figure 4 but using the nu-clear L IR of galaxies (scaled as the far-IR flux; see § µ m monochromatic luminosity, and it is color coded asa function of the 6.2 µ m PAH EW. Only pure star-forming LIRGs,defined as to have 6.2 µ m PAH ≥ . µ m, are shown. The solid lineis a fit to the data. See Eq. (4). compact galaxies that is usually associated with the mostrecent starburst. In Figure 5 we fit the data to providea relation between the [C ii ]/FIR ratio and the Σ IR forpure star-forming LIRGs. The analytic expression of thefit is: log ( [C ii ]FIR ) = 0 . ± . − . ± . × log (Σ IR )(4)with a dispersion in the y-axis of 0.16 dex. The slopeand intercept of this trend are indistinguishable (withinthe uncertainties) from those obtained in Eq. (3), whichwas derived by fitting all data-points including low6.2 µ m PAH EW sources with measured mid-IR sizes.This further supports the idea that the influence of AGNactivity is negligible among IR-selected galaxies with10 − < [C ii ]/FIR < − and that the increase in IRluminosity of these sources is due to a boost of theirwarm dust emission. The Influence of AGN in the [C ii ] Deficit Figure 6 shows the [C ii ]157.7 µ m/FIR ratio as a func-tion of the 6.2 µ m PAH EW for the LIRGs in GOALS.Starburst sources with large PAH EWs have a mean[C ii ]/FIR ratio of 4 . × − with a standard deviationof 2 . × − . As the 6.2 µ m PAH EW becomes smallerthe dispersion increases and we find galaxies with bothvery small ratios as well as sources with normal values (orslightly lower than those) typical of purely star-formingsources (see also Sargsyan et al. 2012).If AGNs contribute significantly to the far-IR emis-sion of LIRGs and/or suppress PAH emission via photo-evaporation of its carriers, we would expect AGN-dominated sources to show significantly low [C ii ]/FIRratios and small PAH EWs. We have used the Spitzer /IRS spectra of our galaxies to identify which0 D´ıaz-Santos et al.
Fig. 6.— [C ii ]157.7 µ m/FIR ratio for individual galaxies in theGOALS sample as a function of the 6.2 µ m PAH EW measuredwith Spitzer /IRS. Galaxies are color-coded as a function of their νL ν µ m ratio. If this information is not available, galax-ies are shown as small black circles. The solid line representsthe range in [C ii ]/FIR and 6.2 µ m PAH with increasing contri-bution from an AGN (see text for details). The dotted lines are × × . ii ]/6.2 µ m PAH ratio and the L FIR / νL ν µ m relation forstar-forming galaxies. The dashed line assumes a decreasing of the[C ii ]/6.2 µ m PAH ratio proportional to the 6.2 µ m PAH EW dueto pure PAH destruction from an AGN. of them are hosting an AGN based on several mid-IR diagnostics: (1) [Ne v ]14.32 µ m/[Ne ii ]12.81 µ m > .
5; (2) [O iv ]25.89 µ m/[Ne ii ]12.81 µ m >
1; (3) S ν µ m/ S ν µ m < α > − . S ν ∝ ν α );as well as (4) the 6.2 µ m PAH EW itself (see constraintsabove). All these thresholds are rather restrictive andensure that the contribution of an AGN to the mid-IRluminosity of a galaxy is at least − µ m PAH EWs and noother AGN signatures, and is a more conservative cutthan applied in Petric et al. (2011) to identify potentialAGNs. Strikingly, these sources are not preferentiallyfound at the bottom-left of the parameter space but in-stead as many as 2/3 show [C ii ]/FIR > − , typical ofstar-forming sources with large 6.2 µ m PAH EWs. Thissuggests that the impact of the AGN on the far-IR lu-minosity of these mid-IR dominated AGN LIRGs is verylimited, unless the it contributes to both the [C ii ] andfar-IR in the same relative amount as the starburst does.While ∼
18% of our sample appears to have signifi-cant AGN contribution to the mid-IR emission (Petricet al. 2011), the fraction in which the AGN dominatesthe bolometric luminosity of the galaxy is much smaller.To investigate this, we use two of the indicators describedabove, the [O iv ]25.89 µ m line and the 6.2 µ m PAH EW,and the formulation given in Veilleux et al. (2009) to cal-culate the bolometric AGN fraction of those galaxies withat least two mid-IR AGN detections. We find that onlyfour (20%) of these galaxies have contributions >
50 %in both indicators (black crosses in Figure 6). Two galax-ies have [C ii ]/FIR < − (33%) and two (14%) a largerratio. To quantitatively asses the relationship between AGNactivity and the [C ii ]/FIR ratio among galaxies host-ing an AGN, it is important to estimate first the AGNcontribution to the far-IR flux. If we assume that theratio of [C ii ]157.7 µ m to 6.2 µ m PAH emission of thestar-forming LIRGs in GOALS is constant, as is thecase for most normal, lower luminosity galaxies (Helouet al. 2001; Croxall et al. 2012; Beir˜ao et al. 2012), wecan calculate the expected evolution of the [C ii ]/FIRratio as a function of the 6.2 µ m PAH EW if we alsoassume that pure starbursts have a typical 6.2 µ m PAHEW SB = 0.65 µ m and [C ii ]/FIR = 4 . × − (as shownabove), and that the average L FIR / νL ν µ m ratios forpure star-forming galaxies and AGNs are ∼
15 and ∼ ∼ −
25 in our sample while the value for AGNshas been estimated from the intrinsic AGN SED of Mul-laney et al. (2011). The predicted trend is shown in Fig-ure 6 as a solid black line, which agrees very well withthe location of the AGNs identified with at least twomid-IR indicators (red stars). Under these assumptions,the 6.2 µ m PAH EW has to be reduced by a factor of ∼
15 with respect to the 6.2 µ m PAH EW SB , i.e., downto ≃ . µ m, before the AGN can contribute 50 % to the L FIR . In fact, 2/3 of galaxies with EWs lower than thisthreshold have been identified as harboring an AGN bytwo or more mid-IR diagnostics. Therefore, only whenthe AGN contribution to the far-IR flux is significant dowe see a noticeable decrease of the [C ii ]/FIR ratio (al-ways < − ). We note however that the contrary mightnot be necessarily true since there are galaxies with low[C ii ]/FIR ratios but with 6.2 µ m PAH EWs & . µ m.We emphasize that this prediction does not account forpossible destruction of PAH molecules due to the AGN.However, if the reduction of the PAH EW was entirelydue to this effect, we would expect a linear correlationbetween the [C ii ]/FIR ratio and the 6.2 µ m PAH EW,which is described by the dashed line in Figure 6. Aswe can see, the mid-IR AGN dominated galaxies do notfollow the predicted trend, suggesting that PAH destruc-tion is not important in LIRGs with [C ii ]/FIR & − at least at the scales probed by Herschel and
Spitzer , inagreement with the results obtained in D´ıaz-Santos et al.(2011).Nearly half of galaxies with [C ii ]/FIR < − and6.2 µ m PAH EW < . µ m have no other direct mid-IR diagnostic that reveals the presence of an AGN. In-terestingly, all of them are among the outliers found inFigure 3, showing an excess in the S . µ m with respect totheir observed [C ii ]/FIR. We argued in § ii ] and far-IR. The energy source of thiscomponent is unknown, though, since both an AGN oran ultra compact H ii region could generate such mid-IRsignatures. However, the monochromatic νL ν µ mratios displayed by these objects are & ∼ − µ m, with νL ν = constant, and fading be-xplaining the [C ii ] Deficit in LIRGs – First GOALS Results From Herschel/PACS 11yond. This adds evidence to the result obtained abovethat this type of deeply embedded objects only dominatethe luminosity of the galaxy in the mid-IR.Furthermore, we would like to emphasize that the factthat the source of this warm, compact emission does notproduce the detected PAH or [C ii ] emissions rules outmodels where PAH obscuration is invoked to explain thelow PAH EWs found in these sources, since their ob-served [C ii ] flux compared to that of the far-IR is alsovery low, implying that it is not extinction but rather thefact that the PDR emission of the warmest dust compo-nent in these LIRGs is actually extremely limited. IMPLICATIONS FOR INTERMEDIATE ANDHIGH-REDSHIFT GALAXY SURVEYS
At intermediate redshifts, z ∼ −
3, it has been foundthat IR-luminous galaxies span a wide range in [C ii ]/FIRratios: ∼ − − − . (Stacey et al. 2010). A sur-prising discovery came from the most luminous systems,and the fact that many of them show values of this ratiosimilar to those found in local, lower luminosity galax-ies (e.g., Maiolino et al. 2009; Hailey-Dunsheath et al.2010; Sturm et al. 2010; Stacey et al. 2010). Theseresults, added to a number of recent findings obtainedfrom the analysis of mid-IR dust features of star-forminggalaxies using Spitzer /IRS spectroscopy (e.g., Pope et al.2008; Murphy et al. 2009; Desai et al. 2009; Men´endez-Delmestre et al. 2009; D´ıaz-Santos et al. 2010b, 2011;Rujopakarn et al. 2011; Stierwalt et al. 2013a) arepointing towards an emerging picture in which the lo-cal counterparts of the dominant population of IR-brightgalaxies at intermediate and high redshifts ( z >
1) arenot extremely dusty systems with similar IR luminosities(i.e., local ULIRGs) but rather galaxies with more mod-est SFRs, or L IR ≃ − L ⊙ (starbursts and LIRGs).Therefore, since GOALS is a complete, flux-limited sam-ple of 60 µ m rest-frame selected LIRGs systems in thelocal Universe covering an IR luminosity range from ∼ to ∼ L ⊙ , the empirical relations we findcan be used to estimate what might be seen in similarsurveys of intermediate and high redshift IR-luminousgalaxies.Recently, Elbaz et al. (2011) has found that the ma-jority of star-forming galaxies, from the nearby Universeand up to z ∼
2, follow a ”main sequence” (MS) that isdepicted by a specific SFR (SSFR MS ) that increases withredshift. Galaxies in the MS are characterized by pro-ducing stars in a quiescent mode, with very low efficien-cies ( L IR / M gas . L ⊙ / M ⊙ ) and extended over spatialscales of several kpc (Daddi et al. 2010a; Magdis et al.2012). On the other hand, galaxies with high SSFRsare currently experiencing a strong and very efficient,but short-lived (less than few hundred Myr) starburstevent, some of them probably consequence of a majormerge interaction. The SSFR of local galaxies is anti-correlated with their compactness as measured in themid-IR, as well as from radio wavelengths (see also Mur-phy et al. 2013). Therefore, the trends of [C ii ]/FIR withFEE . µm , Σ µ m , and Σ IR found in § § ii ]/FIR ratios should belongto the MS while galaxies with low ratios will likely becompact, starbursting sources. Because > −
90% of the UV and optical lightof star-forming LIRGs and ULIRGs is reprocessed bydust into the IR wavelengths (Howell et al. 2010), the L IR / M ⋆ ratio is equivalent to their SSFR since theSFR is directly proportional to the IR luminosity, withSFR IR / L IR = 1.72 × − M ⊙ yr − L ⊙− (as derived inKennicutt 1998). Using Eq. (13) from Elbaz et al. (2011)we calculate that the SSFR of MS galaxies in the lo-cal Universe is SSFR MS ≃ .
09 Gyr − . However, thisequation does not take into account the dependence ofthe SSFR MS on the stellar mass of galaxies. Thus, wecombine it with the SFR vs. M ⋆ correlation obtainedfrom the SDSS sample by Elbaz et al. (2007), using apower-law index of 0.8 for the dependence of the SFRon M ⋆ (their Eq. (5)) and after normalizing it by theSSFR MS at z = 0. This joint equation assumes thatthe exponential dependence of the SFR MS as a functionof M ⋆ does not vary with z , which is roughly the caseat least up to z ∼ − MS ). If we define starbursting galaxies asthose having a SSFR > × SSFR MS , then ∼
68 % of thegalaxies in GOALS would be classified as such.Figure 7 shows the [C ii ]/FIR ratio as a function ofthe integrated SSFR normalized to the representativeSSFR MS of galaxies at z ∼ µ m IRAC luminosities using the M ⋆ /L conver-sions from Lacey et al. (2008). The solid line representsa fit to pure star-forming galaxies with 6.2 µ m PAH EWs ≥ . µ m. The Pearson’s test yields r = − .
65 ( p r = 0),while the Kendall’s test provides κ = − .
47. The corre-lation coefficient derived from the robust fit is − .
76. Wenote that the SFR and M ⋆ plotted here represent inte-grated measurements of our galaxies while the [C ii ] andFIR values were obtained from a single PACS spaxel,probing an area of ∼ ′′ × ′′ , which at the mediandistance of our LIRGs is equal to a projected physicalsize of ∼ ′′ beam at z ∼ × ii ]/FIR and normalized SSFR isnot surprising since the IR luminosity appears in bothquantities. Nevertheless, the correlation is indeed prac-tical in terms of its predictive power. For example, wecan see that there are no star-forming MS galaxies with[C ii ]/FIR < − . Their median ratio is 4 . × − .On the other hand, starbursting sources show a largerrange of ratios, from 10 − to 10 − , with a median of1 . × − . The correlation between the [C ii ]/FIR andthe SSFR/SSFR MS is given by the following equation: log ( [C ii ]FIR ) = − . ± . − . ± . × log ( SSF RSSF R MS )(5)with a dispersion in the y-axis of 0.26 dex. If the sepa-ration between MS and starbursting galaxies (high/low[C ii ]/FIR) at any given redshift is related to an increaseof the star formation efficiency (SFE; higher L IR / M H ;2 D´ıaz-Santos et al. Fig. 7.— [C ii ]157.7 µ m/FIR ratio versus integrated SSFR nor-malized to the SSFR MS ( z ∼
0) for individual galaxies in theGOALS sample. Galaxies are color-coded as a function of their6.2 µ m PAH EW. Colored circles indicate sources for which anEW is available. Squares indicate lower limits. Small black cir-cles are sources for which there is no information. The solid line isa fit to pure star-forming LIRGs only, excluding those sources thatmay harbor an AGN that could dominate their mid-IR emission(6.2 µ m PAH EWs < . µ m; see Figure 6). The dotted lines arethe ± σ uncertainty. The dashed line indicates the SSFR of MSgalaxies at any redshift or M ⋆ . The dotted-dashed line represents3 × SSFR MS , the limit above which galaxies are considered to bestarbursting. The open squares represent two intermediate redshiftgalaxies at z ∼ .
2, while the open diamonds represent three highredshift sub-millimeter galaxies at z ∼ . see, e.g., Graci´a-Carpio et al. 2011; Sargent et al. 2013)then this equation can be applied to predict the [C ii ]luminosity of star-forming galaxies at any z for which ameasurement of the (far-)IR luminosity and stellar massare available as long as their SSFR is normalized to theSSFR MS at that z to account for the increase in gas massfraction ( M gas / M ⋆ ) and therefore SSFR of MS galaxiesat higher redshifts (Daddi et al. 2010a; Magdis et al.2012). Conversely, in future large IR surveys like thoseprojected with the Cornell-Caltech Atacama Telescope(CCAT), this relation could be used for estimating theSSFR or stellar mass of detected galaxies when their [C ii ]and far-IR luminosities are known.As an example, in Figure 7 we show two interme-diate redshift z ∼ . z ∼ . L IR ≃ × L FIR . Since Eq. (13) from Elbaz et al. (2011)is calibrated only up to z ∼ MS on M ⋆ is also uncertain beyond this redshift,we normalized the SSFR of the SMGs by the SSFR MS at z = 3. As we can see, three galaxies follow the corre-lation suggesting that they are starbursting sources with[C ii ]/FIR ratios consistent with their normalized SS-FRs. On the other hand, the two remaining high- z SMGsshow significantly lower [C ii ]/FIR ratios than the aver-age of LIRGs for the same normalized SSFR. In particu-lar, one of them display a [C ii ]/FIR more than an order of magnitude lower than the value predicted by the fitto our local galaxy sample. Interestingly, both SMGs liein the parameter space where most of the mid-IR identi-fied AGN are located (red stars). This may suggest thatthese two galaxies could harbor AGN or unusually week[C ii ] emission for their normalized SSFR.Eq. (3) and (4) can also be used to predict the mid-and total IR luminosity surface density of star-forminggalaxies at high redshifts for which the [C ii ] and far-IRfluxes are known, such as those that may be found infuture spectroscopic surveys with the X-Spec instrumenton CCAT. With instantaneous coverage over all the at-mospheric windows between 190 and 440 GHz, X-Specwill access the [C ii ] line at redshifts from ∼ . ∼ & Herschel in deep fields.In this cases, the predicted mid-IR size of galaxies couldbe compared with direct measurements of the size of theirfar-IR emitting region as observed with ALMA on phys-ical scales similar to those we are probing in our GOALSLIRGs with PACS.Finally, because GOALS is a complete flux-limitedsample of local LIRGs, we are able to predict the con-tamination of sources hosting AGNs in future large-scalesurveys with both [C ii ] and far-IR measurements. InTable 2 we provide the percentages of galaxies with mid-IR detected AGNs classified in different [C ii ]/FIR and S ν µ m/ S ν µ m bins. The values provided in theTable were computed using two conditions for the de-tection of the AGN that serve as upper and lower lim-its for the estimated fractions (columns (2) and (3)).The first was based on the 6.2 µ m PAH EW only andthe second required of an additional mid-IR diagnos-tic to classify the galaxy as harboring an AGN (seeabove). For example, we predict an AGN contamination(based on the 6.2 µ m PAH EW only) of up to ∼
70% for[C ii ]/FIR < × − , which implies that at least ∼ ii ]/FIR ra-tios will be powered by starbursts. Moreover, at the lev-els of [C ii ]/FIR ≥ × − or 63/158 µ m <
2, the AGNdetection fraction is expected to be . − CONCLUSIONS
We obtained new
Herschel /PACS [C ii ]157.7 µ m spec-troscopy for 200 LIRG systems in GOALS, a 60 µ m flux-limited sample of all LIRGs detected in the nearby Uni-verse. A total of 241 individual galaxies where observedin the [C ii ]157.7 µ m line. We combined this informationtogether with Spitzer /IRS spectroscopic data to providethe context in which the observed [C ii ] luminosities and[C ii ]/FIR ratios are best explained. We have found thefollowing results: • The LIRGs in GOALS span two orders of magni-tude in [C ii ]/FIR, from ∼ − to 10 − , with a me-dian of 2 . × − . ULIRGs have a median of 6 . × − . The L [C II] range from ∼ to 2 × L ⊙ for the whole sample. The [C ii ]/FIR ratio is cor-related with the far-IR S ν µ m/ S ν µ m con-tinuum color. We find that all galaxies follow thesame trend independently of their L IR , suggestingxplaining the [C ii ] Deficit in LIRGs – First GOALS Results From Herschel/PACS 13 TABLE 2
Fraction of AGN [C ii ]/FIR AGN-frac AGN-frac 63/158 µ mrange 6.2 µ m PAH multi median(1) (2) (3) (4) > × −
4% 2% 0.52(1 . − × −
15% 10% 0.93(0 . − . × −
18% 9% 1.22 < × −
72% 22% 1.9263/158 µ m AGN-frac AGN-frac [C ii ]/FIRrange 6.2 µ m PAH multi median(1) (2) (3) (4) < . × − . − × − − × − > × − Note . — Top: (1) Range of [C ii ]/FIR ratio; (2) Percentage ofmid-IR detected AGN based only on the 6.2 µ m PAH EW of galaxies( < . µ m; see § S ν µ m/ S ν µ m continuum flux density ratio within the range of[C ii ]/FIR given in column (1).Bottom: (1) Range of S ν µ m/ S ν µ m continuum flux densityratio; (2) Percentage of mid-IR detected AGN based only on the6.2 µ m PAH EW of galaxies (see text); (3) Percentage of mid-IR de-tected AGN based on multiple line and continuum emission diag-nostics (see text); (4) Median [C ii ]/FIR ratio within the range of S ν µ m/ S ν µ m given in column (1). that the main observable linked to the variation ofthe [C ii ]/FIR ratio is the average dust tempera-ture of galaxies, which is driven by an increase ofthe ionization parameter, < U > . • There is a clear trend for LIRGs with deeper 9.7 µ msilicate strengths ( S . µ m ), higher mid-IR luminos-ity surface densities (Σ MIR ), smaller fractions ofextended emission (FEE . µm ) and higher SSFRsto display lower [C ii ]/FIR ratios. These correla-tions imply the the dust responsible for the mid-IR absorption must be directly linked to the pro-cess driving the observed [C ii ] deficit. LIRGs withlower [C ii ]/FIR ratios are more warm and com-pact (higher mid- and IR luminosity surface den-sities, Σ (M)IR ), regardless of what is the origin ofthe nuclear power source. However, this trend isclearly seen also among pure star-forming LIRGsonly, implying that it is the compactness of thestarburst, and not AGN activity as identified in themid-IR, that is the main driver for the declining ofthe [C ii ] to far-IR dust emission. This implies thatthe [C ii ] luminosity is not a good indicator of theSFR in LIRGs with high T dust or large Σ IR sinceit does not scale linearly with the warm dust emis-sion most likely associated to the youngest stars.There are a small number of LIRGs that have alarger [C ii ]/FIR ratio than suggested by their deep S . µ m and warm dust emission. The origin of theenergy source of these LIRGs is unknown, althoughthey likely contain a deeply buried, compact sourcewith little or no PDR emission. • Pure star-forming LIRGs (6.2 µ m PAH EW ≥ . ii ]/FIR = 4 . × − with a stan-dard deviation of 2 . × − , while galaxies withlow 6.2 µ m PAH EWs span the entire range in[C ii ]/FIR. A significant fraction (70 %) of the LIRGs in which an AGN is detected in the mid-IR have [C ii ]/FIR ratios ≥ − , similar to thoseof starburst galaxies suggesting that most AGNsdo not contribute substantially to the far-IR emis-sion. Thus, only in the most extreme cases when[C ii ]/FIR < − might the AGN contribution besignificant. • The completeness of the GOALS LIRG samplehas allowed us to provide meaningful predictionsabout the [C ii ], Σ MIR , and AGN contamination oflarge samples of IR-luminous high-redshift galax-ies soon to be observed by ALMA or CCAT. Ina far-IR selected survey of high- z LIRGs we ex-pect to find up to ∼
70% of AGN contaminationfor [C ii ]/FIR < × − , which implies that atleast 1/3 of IR-selected sources with extremely low[C ii ]/FIR will be powered by starbursts. More-over, above this ratio the AGN fraction is expectedto be . − ii ] andfar-IR emission measurements we can predict theIR luminosity surface density of galaxies, whichcould be compared with direct measurements ofthe size of their far-IR emitting region as observedwith ALMA on physical scales similar to those weare probing in our GOALS LIRGs with PACS. ACKNOWLEDGMENTS
We thank the referee for his/her useful comments andsuggestions which significantly improved the qualityof this paper. We also thank David Elbaz, AlexanderKarim, J. D. Smith, Moshe Elitzur, and J. Gracia-Carpiofor very fruitful discussions. L. A. acknowledges thehospitality of the Aspen Center for Physics, which issupported by the National Science Foundation GrantNo. PHY-1066293. V. C. would like to acknowledgepartial support from the EU FP7 Grant PIRSES-GA-2012-31578. This work is based on observations madewith the
Herschel Space Observatory , an EuropeanSpace Agency Cornerstone Mission with science instru-ments provided by European-led Principal Investigatorconsortia and significant participation from NASA. The
Spitzer Space Telescope is operated by the Jet Propul-sion Laboratory, California Institute of Technology,under NASA contract 1407. This research has madeuse of the NASA/IPAC Extragalactic Database (NED),which is operated by the Jet Propulsion Laboratory,California Institute of Technology, under contract withthe National Aeronautics and Space Administration, andof NASA’s Astrophysics Data System (ADS) abstractservice. APPENDIX In § ii ]157.7 µ m/FIR ra-tio as a function of the PACS-based S ν µ m/ S ν µ mratio for the galaxies in the GOALS sample observed with Herschel , and explain the reasons for adopting and plot-ting this far-IR color instead the more commonly used
IRAS -based S ν µ m/ S ν µ m color. Here we show acomparison between both, to provide the reader with atool for interpreting our results in terms of IRAS colors,if necessary. Figure 8 shows that the far-IR ratios corre-late well ( r = 0 . p r = 0), as expected, with a slope4 D´ıaz-Santos et al. Fig. 8.—
PACS-based S ν µ m/ S ν µ m ratio as a functionof IRAS -based S ν µ m/ S ν µ m ratio for individual galaxies inthe GOALS sample. In LIRG systems with two or more galaxies,the same IRAS color is plotted for each individual galaxy. Thecalculation of the L FIR and the symbols are as in Figure 2. of 1.80 ± ± µ m ratios but the same 60/100 µ m ratio, as theycorrespond to a single, unresolved IRAS source. Never-theless, independently of which of these far-IR colors weutilize, the same overall trend seen in Figure 2 emerges,namely warmer galaxies display smaller [C ii ]/FIR ra-tios. As mentioned in § IRAS far-IR colors of integrated systems, our LIRG samplefollows the same trend found for normal and moderateIR-luminous galaxies observed by
ISO . However, whenusing the 63/158 µ m ratio, the anti-correlation is tighterthan that obtained when the emission from entire sys-tems is employed, likely due to the fact that we are ableto disentangle the true far-IR colors of individual galax-ies. Moreover, the observed 63/158 µ m continuum ratiosspan a much larger dynamical range (a factor of ∼
10, incontrast with the factor of ∼ µ m ratios), which translates in a more accuratesampling of the average dust temperature of LIRGs. REFERENCESAbel, N. P., Dudley, C., Fischer, J., Satyapal, S., & van Hoof,P. A. M. 2009, ApJ, 701, 1147Alonso-Herrero, A., Pereira-Santaella, M., Rieke, G. H., &Rigopoulou, D. 2012, ApJ, 744, 2Armus, L., et al. 2007, ApJ, 656, 148—. 2009, PASP, 121, 559Beir˜ao, P., et al. 2012, ApJ, 751, 144Berta, S., et al. 2011, A&A, 532, A49Brauher, J. R., Dale, D. A., & Helou, G. 2008, ApJS, 178, 280Caputi, K. I., et al. 2007, ApJ, 660, 97Colbert, J. W., et al. 1999, ApJ, 511, 721Contini, M. 2012, MNRAS, 247Croxall, K. V., et al. 2012, ApJ, 747, 81da Cunha, E., Charmandaris, V., D´ıaz-Santos, T., Armus, L.,Marshall, J. A., & Elbaz, D. 2010, A&A, 523, A78Daddi, E., et al. 2010a, ApJ, 713, 686De Breuck, C., Maiolino, R., Caselli, P., Coppin, K.,Hailey-Dunsheath, S., & Nagao, T. 2011, A&A, 530, L8Desai, V., et al. 2007, ApJ, 669, 810—. 2009, ApJ, 700, 1190D´ıaz-Santos, T., et al. 2010b, ApJ, 723, 993—. 2011, ApJ, 741, 32Draine, B. T. 1978, ApJS, 36, 595Elbaz, D., et al. 2007, A&A, 468, 33—. 2011, A&A, 533, A119Elvis, M., et al. 1994, ApJS, 95, 1Fischer, J., et al. 2010, A&A, 518, L41Genzel, R., et al. 1998, ApJ, 498, 579Gonz´alez-Alfonso, E., Smith, H. A., Fischer, J., & Cernicharo, J.2004, ApJ, 613, 247Goulding, A. D., Alexander, D. M., Bauer, F. E., Forman, W. R.,Hickox, R. C., Jones, C., Mullaney, J. R., & Trichas, M. 2012,ApJ, 755, 5Graci´a-Carpio, J., et al. 2011, ApJ, 728, L7Haan, S., et al. 2011, AJ, 141, 100Habing, H. J. 1968, Bull. Astron. Inst. Netherlands, 19, 421Hailey-Dunsheath, S., Nikola, T., Stacey, G. J., Oberst, T. E.,Parshley, S. C., Benford, D. J., Staguhn, J. G., & Tucker, C. E.2010, ApJ, 714, L162Hayes, M. A., & Nussbaumer, H. 1984, A&A, 134, 193Helou, G., Khan, I. R., Malek, L., & Boehmer, L. 1988, ApJS, 68,151Helou, G., Malhotra, S., Hollenbach, D. J., Dale, D. A., &Contursi, A. 2001, ApJ, 548, L73Howell, J. H., et al. 2010, ApJ, 715, 572Iwasawa, K., et al. 2011, A&A, 529, A106 Karim, A., et al. 2011, ApJ, 730, 61Kaufman, M. J., Wolfire, M. G., Hollenbach, D. J., & Luhman,M. L. 1999, ApJ, 527, 795Kennicutt, Jr., R. C. 1998, ApJ, 498, 541Lacey, C. G., Baugh, C. M., Frenk, C. S., Silva, L., Granato,G. L., & Bressan, A. 2008, MNRAS, 385, 1155Laurent, O., Mirabel, I. F., Charmandaris, V., Gallais, P.,Madden, S. C., Sauvage, M., Vigroux, L., & Cesarsky, C. 2000,A&A, 359, 887Levenson, N. A., Sirocky, M. M., Hao, L., Spoon, H. W. W.,Marshall, J. A., Elitzur, M., & Houck, J. R. 2007, ApJ, 654,L45Luhman, M. L., et al. 1998, ApJ, 504, L11Magdis, G. E., et al. 2012, ApJ, 760, 6Magnelli, B., Elbaz, D., Chary, R. R., Dickinson, M., Le Borgne,D., Frayer, D. T., & Willmer, C. N. A. 2011, A&A, 528, A35+Maiolino, R., Caselli, P., Nagao, T., Walmsley, M., De Breuck,C., & Meneghetti, M. 2009, A&A, 500, L1Malhotra, S., et al. 1997, ApJ, 491, L27—. 2001, ApJ, 561, 766Men´endez-Delmestre, K., et al. 2009, ApJ, 699, 667Mullaney, J. R., Alexander, D. M., Goulding, A. D., & Hickox,R. C. 2011, MNRAS, 414, 1082Murphy, E. J., Chary, R., Alexander, D. M., Dickinson, M.,Magnelli, B., Morrison, G., Pope, A., & Teplitz, H. I. 2009,ApJ, 698, 1380Murphy, E. J., Chary, R.-R., Dickinson, M., Pope, A., Frayer,D. T., & Lin, L. 2011, ApJ, 732, 126Negishi, T., Onaka, T., Chan, K.-W., & Roellig, T. L. 2001,A&A, 375, 566Nenkova, M., Sirocky, M. M., Nikutta, R., Ivezi´c, ˇZ., & Elitzur,M. 2008b, ApJ, 685, 160Nikola, T., Genzel, R., Herrmann, F., Madden, S. C., Poglitsch,A., Geis, N., Townes, C. H., & Stacey, G. J. 1998, ApJ, 504, 749Paladini, R., et al. 2012, ApJ, 760, 149Petric, A. O., et al. 2011, ApJ, 730, 28Pilbratt, G. L., et al. 2010, A&A, 518, L1Poglitsch, A., et al. 2010, A&A, 518, L2Pope, A., et al. 2008, ApJ, 675, 1171Rujopakarn, W., Rieke, G. H., Eisenstein, D. J., & Juneau, S.2011, ApJ, 726, 93Sanders, D. B., & Mirabel, I. F. 1996, ARA&A, 34, 749Sargent, M. T., et al. 2013, ArXiv e-printsSargsyan, L., et al. 2012, ArXiv e-printsSiebenmorgen, R., Kr¨ugel, E., & Spoon, H. W. W. 2004, A&A,414, 123 xplaining the [C ii ] Deficit in LIRGs – First GOALS Results From Herschel/PACS 15] Deficit in LIRGs – First GOALS Results From Herschel/PACS 15