A Survey of HC3N in Extragalactic Sources - Is HC3N a Tracer of Activity in ULIRGs?

0.71 MB·English

by  J. E. Lindberg

Astronomy&Astrophysicsmanuscriptno.15565 ©ESO2011
January11,2011
A Survey of HC N in Extragalactic Sources
3
Is HC N a Tracer of Activity in ULIRGs?
3
J.E.Lindberg1,2,3⋆,S.Aalto3⋆⋆,F.Costagliola3⋆⋆⋆,J.-P.Pe´rez-Beaupuits4,5,R.Monje6,andS.Muller3
1 Centre for Star and Planet Formation, Natural History Museum of Denmark, University of Copenhagen, Øster Voldgade 5-7,
DK-1350CopenhagenK,Denmarke-mail:[email protected]
2 NordicOpticalTelescope,Apartado474,E-38700SantaCruzdeLaPalma,SantaCruzdeTenerife,Spain
3 DepartmentofEarthandSpaceSciences,OnsalaObservatory,ChalmersUniversityofTechnology,SE-43992Onsala,Sweden
1 4 Max-Planck-Institutfu¨rRadioastronomie,AufdemHu¨gel69,53121Bonn,Germany
1 5 KapteynAstronomicalInstitute,UniversityofGroningen,Landleven12,9747ADGroningen,TheNetherlands
0 6 CaliforniaInstituteofTechnology,1200E.CaliforniaBlvd.,MailCode301-17,Pasadena,CA91125-4700,USA
2

n
ReceivedAugust11,2010;acceptedDecember16,2010
a
J
ABSTRACT
0
1
Context.HC NisamoleculethatismainlyassociatedwithGalacticstar-formingregions,butithasalsobeendetectedinextragalactic
3
environments.
]
Aims.Topresent thefirstextragalacticsurvey of HC N, whencombining earlier datafromtheliteraturewithsixnew single-dish
O 3
detections,andtocompareHC Nwithothermoleculartracers(HCN,HNC),aswellasotherproperties(silicateabsorptionstrength,
3
C IRfluxdensityratios,Ciiflux,andmegamaseractivity).
h. Methods.WepresentmmIRAM30m,OSO20m,andSESTobservationsofHC3Nrotationallines(mainlytheJ=10-9transition)
andoftheJ=1-0transitionsofHCNandHNC.OurcombinedHC Ndataaccountfor13galaxies(excludingtheupperlimitsreported
p 3
forthenon-detections),whilewehaveHCNandHNCdataformorethan20galaxies.
o- Results.Apreliminarydefinition“HC3N-luminousgalaxy”ismadebasedupontheHC3N/HCNratio.Most(∼80%)HC3N-luminous
galaxiesseem tobe deeply obscured galaxiesand (U)LIRGs.A majority(∼60%or more) of the HC N-luminous galaxies inthe
r 3
t samplepresentOHmega-orstrongkilomaser activity.ApossibleexplanationisthatbothHC NandOHmegamasersneedwarm
s 3
dustfortheirexcitation.Alternatively,thedustthatexcitestheOHmegamaseroffersprotectionagainstUVdestructionofHC N.A
a 3
[ highsilicateabsorptionstrengthisalsofoundinseveraloftheHC3N-luminousobjects,whichmayhelptheHC3Ntosurvive.Finally,
wefindthatahighHC N/HCNratioisrelatedtoahighdusttemperatureandalowCiiflux.
3
1
v Key words. galaxies: ISM – galaxies: starburst – galaxies: active – radio lines: galaxies – radio lines: ISM – ISM: molecules –
1 molecules:HC3N,HCN,HNC
5
7
1 1. Introduction vated HCN/HCO+ line ratio (under the circumstances that the
1. line ratio directly reflects the abundanceratio). However,more
0 Finding useful tracers of the interaction between the activity recentchemicalmodelsinsteadsuggestthatHCO+ isenhanced
in galaxy nuclei and surrounding interstellar medium (ISM)
1 inXDRs(Meijerink&Spaans2005;Meijerinketal.2007),and
1 is an important and growing aspect of current extragalactic HCO+ isalsoexpectedtobeunder-abundantinregionsofvery
molecularastronomy.Inthiscontext,singledishsurveysofpo-
: youngstar formation(Aalto 2008), so the line ratio is ambigu-
v lar molecules such as HCN, HNC, HCO+, and CS have been
ous.
i
X usedtoinvestigatepossiblecorrelationsbetweenmolecularline
Other molecular tracers could help resolve the dichotomy
ratios and type/intensity of activity (e.g. Kohnoetal. 2001;
r of the HCN/HCO+ line ratio. The serendipitous discovery of
a Aaltoetal. 2002; Imanishietal. 2004; Gracia´-Carpioetal.
the J=10-9 line of HC N near the HNC J=1-0 line in a sur-
2006; Kripsetal. 2008; Baanetal. 2008). For example, it has 3
vey by Aaltoetal. (2002) led us to look more closely at this
been suggested that an elevated HCN/HCO+ 1-0 line intensity
molecule. HC N is the simplest of the cyanopolyynes(carbon
ratio indicates the presence of an AGN (Gracia´-Carpioetal. 3
chainswithanattachedCNgroup)andisagrainchemistryprod-
2006). Aroundan active galactic nucleus(AGN) the chemistry
uct, in contrast to molecules such as HCN and HCO+. HC N
is supposedly dominated by hard X-rays in an X-ray domi- 3
thrivesinwarm,denseshieldedregionssuchashotcoreswhere
natedregion(XDR),andsomechemicalmodelspredictanabun-
abundances can reach 10−8 or even higher, since it is easily
dance enhancement of HCN paired with selective destruction
destroyedbyphoto-dissociation(Rodriguez-Francoetal.1998)
of HCO+ (Maloneyetal. 1996) – which could lead to an ele-
and C+ ions (Prasad&Huntress 1980). Therefore, HC N line
3
⋆ J.L. wishes to thank Instrumentcenter for Danish Astrophysics emissioncouldbeusedtoidentifygalaxieswherestarformation
(IDA)forgrantsupport. isintheearly,embeddedstageofitsevolution.Recently,HC3N
⋆⋆ S.A.wishestothanktheSwedishResearchCouncilforgrantsup- was found in high abundance in the highly obscured galaxy
port. NGC 4418 (Aaltoetal. 2007), as well as the ULIRG Arp 220
⋆⋆⋆ F.C.wishestothanktheEUESTRELAprogrammeforsupport. (Aaltoetal.2002).
1
J.E.Lindbergetal.:ASurveyofHC NinExtragalacticSources
3
We have searched for HC N line emission in a sample of the CN to form HC N (Meier&Turner 2005; Chapmanetal.
3 3
galaxies in various stages and types of activity: AGNs, star- 2009;Fukuzawa&Osamura1997):
bursts, and ultraluminous galaxies (ULIRGs). In some of the
galaxiesthenatureoftheactivityiselusivesinceitisembedded C2H2+CN−→HC3N+H. (2)
in huge columns of dust absorbing emission at optical and in-
This hypothesis is strengthened by interferometric maps of
fraredwavelengths.Insomecases,theextinctionissostrongthat
HC NandC HinIC342foundinMeier&Turner(2005).The
no emission emergesat optical or IR wavelengthsrequiring us 3 2
mapsshowa clear anti-correlationbetweenthe distributionsof
toprobethenatureoftheactivityatradioandmmwavelengths.
themolecules.UsingtheabundancesofHC N,C H,andCNin
HC Nhasarichmmandsub-mmwavelengthspectrumconsist- 3 2
3
a region where it is expected that grains with C H first were
ing of a multitude of rotational and vibrational lines often ap- 2 2
present, it should thereforebe possible to tell whether a strong
pearingclosetoeachotherinthesameband.Throughitsvibra-
UVfieldispresentornot.
tionaltransitions, HC N respondsstronglyto the IR field from
3
Irvineetal.(1987)findthatC H,contrarilytoHC N,istwo
dustynuclei(Costagliola&Aalto2010a).Therefore,combining 2 3
orders of magnitude more abundant in the Orion ridge than in
the rotationaland vibrationalline informationof HC N allows
3
its hotcore. Possibly, the hotcore shields the HC N and C H
us to study the abundanceof HC N (comparingwith chemical 3 2 2
3
fromphoto-dissociatingintoC H.SomeoftheC H insteadre-
modelsofXDRsandstarbursts)aswellastheintensityandtem- 2 2 2
acts with the CN (althoughnotveryabundant)and formseven
peraturestructureoftheburiedIRsource.
more HC N. The highest HC N abundances are found in Sgr
Rotational lines of vibrationally excited HC N have re- 3 3
3 B2 hot cores, being in the order of 10−7 of the H abundance
cently been discovered in a few galaxies (NGC 4418 2
(deVicenteetal.2000).
(Costagliola&Aalto 2010a),Arp 220(Mart´ınetal. 2010),and
We have not included all possible HC N formation mech-
IC 860 (Costagliolaetal. 2010b)), therefore showing that it is 3
anisms here, and investigations of other processes are ongo-
important to take both radiative and collisional excitation into
ing, as for example the notion of ice formation of HC N
considerationwheninterpretingHC NlineemissionfromIRlu- 3
3
(Szczepanskietal.2005).
minousgalaxies.
ItcanalsobenotedthatabsorptionlinesofHC Nhasbeen
3
found in a z ∼ 0.89 galaxy located in front of the quasar PKS 2.2.DestructionofHCN
3
1830-211(Henkeletal.2009).
In the Galaxy, HC N is associated with warm, dense, shielded
3
gas around young stars or star-forming regions, and is
1.1.Outline easily destroyed by UV radiation and reactions with C+
ions (Rodriguez-Francoetal. 1998; Meier&Turner 2005). It
Here, the first survey of extragalactic HC N data is presented.
3
will form either C H or C N when being photo-dissociated
WereportnewHC Nobservationsin19galaxies(detectionsin 2 3
3 (Cherchneffetal.1993),andC H+ orC N+ whenreactingwith
sixofthem),mainly(U)LIRGsandstarburstgalaxies,andcom- 3 4
C+ (Bohme&Raksit 1985). Among the possible reactions de-
plete this sample with data fromall earlier extragalacticHC N
3
stroyingHC Nare:
emissionlinesingle-dishdetectionsfoundintheliterature.The 3
aimofthestudyistocomparetheHC Nluminositywithother
3 HC N+hν−→C H+CN, (3)
moleculartracersaswellasgalaxypropertiestoseeifthepres- 3 2
enceofHC3Ncanbeusedtopredictothergalaxyproperties,e.g. HC N+hν−→C N+H, (4)
3 3
thesourceofactivityinthegalaxy.
In Section 2, the general properties of HC3N in space are HC3N+C+ −→C3H++CN, (5)
discussed.InSection3wepresentthenewobservationsanddis-
cuss the collection of data from the literature. In Section 4 we HC N+C+ −→C N++H. (6)
3 4
presenttheresultsintermsoflineintensitiesandlineratios.In
Section5wediscusstheinterpretationoftheHC Nresultsand Reactionratesofthesereactions,aswellasthoseinSection2.1,
3
compare them with silicate absorption strength (Section 5.2), canbefoundine.g.Cherchneffetal.(1993).
OH megamaser activity (Section 5.3), IR flux density ratios
(Section5.4),Ciiflux(Section5.5),andtheHNC/HCN1-0line
2.3.AbundancesofHCN
ratio (Section 5.6). In Section 5.7 future studies resulting from 3
thisprojectarediscussed. Irvineetal. (1987) give the relative abundances of several
molecules in the core and ridge of the Orion molecular cloud.
ThedetectionsofthesemoleculesintheGalaxyareindicativeof
2. HC Ninspace
3 theirabundancesinhighandlowdensitymolecularregions.The
2.1.GenerationofHC3N relativeHC3Nabundanceliesaround10−9 oftheH2 abundance
in the core (dense region), and 10−10 in the ridge (low density
Acetylene, C H , exists on grains in the ISM (Chapmanetal.
2 2 region). This is a relatively small difference between high and
2009). After evaporatingfrom the grains there are at least two lowdensityregions,ascomparedtoe.g.HCN,withabout10−7
different paths the C2H2 may follow. If a high UV field is oftheH abundanceinthecore,and10−9intheridge.
2
present(theregionbeingaPDR),itwillphoto-dissociateintothe
Theintenseradiationfromstarburstregionsand/orAGNsin
ethynyl radical, C H (Meier&Turner 2005; Cherchneffetal.
2 thecentreofmanygalaxieswillturnsurroundinggascloudsinto
1993;Heikkila¨etal.1999):
regionswherethechemicalstructuredependshighlyonthera-
C H +hν−→C H+H. (1) diationfield,eitherphoton-dominatedregions(PDRs)orX-ray
2 2 2
dominatedregions(XDRs). In XDRs the abundanceof several
IfnostrongUVfieldispresent(noPDR),butCN(thecyanorad- molecules(e.g.CNandCH )areexpectedtobeenhancedwith
2
ical) isavailableforreactions,the C H willinsteadreactwith respecttotheabundancecommonlyfoundinPDRs.Duetothe
2 2
2
J.E.Lindbergetal.:ASurveyofHC NinExtragalacticSources
3
highC+abundanceinXDRs,alowHC Nabundanceisexpected Table1.Datafortheobservationsreportedinthiswork
3
(Aalto2008).TheHC N/CNabundanceratiofoundinPDRsis
3
alsoverylowcomparedtothesameratiomeasuredinhot,dense
Galaxy Molecule Line Telescope Date
cores (Rodriguez-Francoetal. 1998). For further discussion
aboutPDR and XDR chemistry,see e.g. Tielens&Hollenbach Arp220 HC N 10-9 OSO 2001-11-11
3
(1985); Lepp&Dalgarno (1996); Meijerink&Spaans (2005); Arp220 HC N 12-11 OSO 2001-11-09
3
Maloneyetal.(1996). Circinus HNC 1-0 SEST 2001-01-14
IC694 HC N 12-11 OSO 2001-11-10
3
IC860 HC N 28-27 IRAM 2007-12-16
3
3. Observations IC860 HNC 1-0 IRAM 2007-12-16
I17208 HC N 10-9 IRAM 2007-12-16
3
The new observations reported in this work were carried out I17208 HNC 1-0 IRAM 2006-06-28
with the IRAM 30 m1, OSO 20 m, and SEST 15 m telescopes Maffei2 HC N 12-11 IRAM 2007-08-25
3
between 2001 and 2008. Detailed lists with system tempera- NGC34 HNC 1-0 SEST 2001-01-14
tures and dates for the observations reported in this work are NGC613 HCN 1-0 SEST 2001-02-11
found in Table 1). The pointing accuracy was better than 2′′ NGC613 HNC 1-0 SEST 2001-02-13
NGC1056 HC N 16-15 IRAM 2006-06-30
for all the observations, and typical system temperatures were 3
NGC1056 HNC 1-0 IRAM 2006-06-30
150 K (IRAM 90-110 GHz), 250 K (IRAM 125 GHz), 400 K
NGC1377 HC N 16-15 IRAM 2007-12-13
(IRAM225-250GHz),and300K(OSOandSEST).Inthissur- 3
NGC1377 HC N 25-24 IRAM 2007-12-13
3
vey we also include data from the literature using the already NGC1377 HCN 1-0 IRAM 2007-12-13
mentionedtelescopes, as well as the NRO 45 m, NRAO 12 m, NGC1377 HNC 1-0 IRAM 2007-12-13
andFCRAO14mtelescopes.Wheneverusingdatafromthelit- NGC1614 HNC 1-0 SEST 2001-02-13
erature, the beam sizes and efficiencies given in the respective NGC2146 HC N 10-9 OSO 2001-11-08
3
articles have been used for calculations of line ratios. The pa- NGC2146 HC3N 12-11 OSO 2001-11-08
rametersusedforallnewobservationsreportedinthisworkare NGC2623 HC3N 12-11 OSO 2001-11-12
NGC3079 HC N 10-9 IRAM 2006-05-14
giveninTable2. 3
NGC3079 HC N 16-15 IRAM 2006-05-14
Onecouldarguethatthemanydifferentinstrumentsusedto 3
NGC3079 HC N 25-24 IRAM 2006-05-14
obtainthe data in this article mightintroducea biasdifficultto 3
NGC3690 HC N 12-11 OSO 2001-11-09
compensatefor. However,when comparingthe HNC/HCN 1-0 3
NGC4418 HCN 1-0 IRAM 2008-07-19
lineratiosobtainedwithIRAM andSESTrespectively,nosys- NGC4945 HNC 1-0 SEST 2001-01-15
tematicbiasisdetected.Theaveragelineratiowascalculatedto NGC5135 HNC 1-0 SEST 2001-01-13
0.51±0.11forSESTdataand0.51±0.08forIRAMdata. NGC6946 HC N 12-11 IRAM 2007-08-25
3
All objects investigated (observed by us or with data from UGC5101 HC N 10-9 IRAM 2007-12-13
3
the literature) are listed in Table 3, along with some important UGC5101 HNC 1-0 IRAM 2007-12-13
characteristics.
The relative HC N abundancescalculated in this work will Notes. Fortheobservationscitedfromtheliterature,seetherespective
3
articlesreferredtoinTables4-6.Wewouldliketopointoutthatsomeof
be expressed as line ratios between an HC N line (mostly the
3
theHC N10-9datareportedbyuscomefromSESTHNC1-0spectra.
J =10-9transition)andtheJ =1-0transitionsofHCNandHNC. 3
Thesetwomoleculesarechosenastheyaregoodtracersofhigh
densityregions(seee.g.Papadopoulos2007;Aaltoetal.2002), Table2.Observationalparameters.
whereweexpecttofindtheHC N(Meier&Turner2005).Also,
3
HCNandHNCdataareavailableformostoftheobjectsinthe Transition ν[GHz]a HPBW[′′]b η b
sample.Wenotethatthelineratiosarenotlinearlyproportional mb
toratiosbetweentheabundancesofthespecies,sincetheywill IRAM:
also depend on excitation conditions and optical depths in the HC3N10-9 90.979 28 0.80
galaxies. A high HC3N/HCN ratio might thus sometimes be a HC3N12-11 109.174 24 0.73
HC N16-15 145.561 17 0.67
tracerofdiscrepanciesintemperatures,densities,orIRpumping 3
HC N25-24 227.419 10.5 0.63
inthegalaxies. 3
HC N28-27 254.699 9 0.59
A discussion of the method used when calculating the line 3
HCN1-0 88.632 28 0.80
ratios can be found in Appendix A. A few of the most nearby HNC1-0 90.664 28 0.80
galaxies in the survey have such a large angular distribution OSO:
that the measured values might not represent a global value HC N12-11 109.174 36 0.52
3
for molecular gas in the galaxy, but rather a value for a cer- HC N10-9 90.979 42 0.59
3
tain (central) region of the galaxy. This effect is discussed in SEST:
AppendixA.1. HC3N10-9 90.979 55 0.75
HCN1-0 88.632 57 0.75
The sample of galaxies observed by us has been chosen to
HNC1-0 90.664 55 0.75
haveahighprobabilityoffindingHC N–itisbynomeansin-
3
tendedtobeanunbiasedsampleofsomerandomgalaxies,and
Notes. (a) From the NIST database Recommended Rest Frequencies
thus the relatively highdetection ratio should definitely notre-
for Observed Interstellar Molecular Microwave Transitions
flecttheamountofHC N-luminousgalaxiesintheuniverse.The
3 (http://physics.nist.gov/cgi-bin/micro/table5/start.pl).
sameismostlikelytrueforthegalaxiesfoundintheliterature.
(b)Thehalf-powerbeamwidthsandmainbeamefficienciesarecollected
fromtherespectivetelescopewebpages.
1 BasedonobservationscarriedoutwiththeIRAM30mTelescope.
IRAM is supported by INSU/CNRS (France), MPG (Germany), and
IGN(Spain)
3
J.E.Lindbergetal.:ASurveyofHC NinExtragalacticSources
3
Anotherimportantselectioneffectfortheobjectsfromtheliter- other HC N line (specified in the footnotes) has been used for
3
atureisthatdetectionsaremuchmorelikelytobereportedthan theratios.
non-detections,whichalsoleadstoabiasedsample. AfewofthegalaxiesappeartwiceinTable7.Forthese,sev-
To increase the chance of detecting HC N, almost all of eralobservationshavebeenfoundforthesametransition.Ascan
3
the galaxies that were chosen to be part of the sample have beseen,thevaluesoftheseobservationsdonotalwaysagree.If
earlier detections of HCN, which means that they should have theHC NandHNCdataarefoundinthesamespectruminone
3
large amounts of dense gas, increasing the possibility of find- of the observations,preference has been given to this observa-
ing HC N. As the goal of the study is to investigate if HC N tion,asitwillincreasetheaccuracyontheHC N/HNCratio.In
3 3 3
cantracethesourceoftheactivityinactivegalaxies,thesample allothercases,thespectrumofeachobservationhasbeeninves-
consists only of active galaxies – starburst galaxies and AGN tigated (when available), and the values from the spectra with
galaxies (the source of the activity is although disputed or un- thelowestnoiselevelshavebeengivenpriorityandareputfirst
knowninmanyofthegalaxiesinthesample). inTable7.
Throughoutthearticle,theT∗ scalewillbeusedforallour
A
data.Forthedatafromtheliterature,thetemperaturescaleused
4.2.1. HC N-luminousgalaxies
ineacharticlewillbeusedinourtables,clearlynotedwhenever 3
theT∗ scaleisnotused.Thiswillmakeiteasiertodetectanyer-
A A definitionofanHC3N-luminousgalaxyisnowdesirable.As
rorsthatmighthaveoccurredinthesurveywork.Whentheline
HCN is the most common dense gas tracer, and also should
ratiosare calculated,the efficiencieswill be taken into account
be a more stable component of the dense gas than HNC, the
properly. HC N/HCN ratios were decided to be used for this definition.
Data analysis was performed with the X-Spec2 software Itse3emslikemostgalaxieshaveHC N/HCNratiosbelow0.15,
3
package.Afirstorderbaselinewassubtractedfromallspectra.
withtheexceptionforafewinterestinggalaxies.Thus,wecon-
siderintherestofthepaperthatgalaxieswith I(HC3N) >0.15are
I(HCN)
4. Results HC N-luminousgalaxies.Ifacorrespondinglimitshouldbeset
3
ontheHC N/HNCratio,itwouldbearound0.25toincludethe
3
All new HC N, HCN, and HNC spectra reported in this work
3 samegalaxies.
aredisplayedinFigures6-12.Theobservedvaluesofthespec-
ThegalaxiesthusseenasHC N-luminousorHC N-richare
3 3
tral line intensities can be found in Tables 4-6. Data from the
NGC4418,IC342,Circinus,M82,Maffei2,Arp220,andIRAS
literaturearealsoincludedinthesetables.
17208-0014.WealsochoosetoincludeIC860,consideringthat
Some observationsof HC N 10-9 and HNC 1-0 performed
3 itsmoderateHC N/HNCratioisfortheHC N28-27transition,
3 3
with SEST include both these lines in the same spectrum, due
asthehighertransitionsseemtobeweakerthantheJ =10-9line
to the large bandwidth. The spectra are in these cases labelled
in most galaxies where more than one line has been observed
accordingtothecentralpeak.Thefrequencydifferencebetween
(whenbeameffectsarecompensatedfor).Afewofthesegalax-
thetwopeaksis315MHz,ascanbeseeninTable2.Thisgives
iesarequitenearby,andasdiscussedabove,theHC N/HCNra-
avelocitydifferenceofapproximately1000km/s. 3
tioofgalaxieswithsourcesizeslargerthanthetelescopebeam
sizewillprobablybeoverestimated.Thisisparticularlythecase
4.1.Newdetections forM82,IC342,andMaffei2.
Some galaxies can definitely be seen as HC N-poor, since
3
Thisis notonlythe firsttextto puttogethera surveyof all ex- theyhaveHC N/HCNratios(orupperlimitsforthisratio)less
3
tragalacticHC N emissionlinedata,butitalsoreportsthefirst
3 thanor equalto 0.10:NGC253,NGC1068,NGC 1808,NGC
HC Ndetectionsinsixgalaxies:Circinus,IC860,IRAS17208-
3 3256,NGC4945,NGC6946,andUGC5101.SinceNGC253
0014,Maffei2,NGC1068,andNGC3079.Thenumberofex-
also belongsto the nearbygalaxies, thisvalue shouldprobably
tragalactic sources where HC N has been detected is thus al-
3 beevenlower.
most doubled. Three of the HNC detections are also made in
sources without earlier HNC detections: Circinus, IC 860, and
IRAS 17208-0014.Finally, the first detection of HCN in NGC
5. Discussion
613isalsoreported.
Itisnotsurprisingtofindthatmostgalaxiesinoursamplefrom
publishedarticlesshowHC Ndetections–otherwisetheywould
4.2.Lineratios 3
notbesubmittedforpublication.However,ifonlycountingthe
Thelineratioshavebeencalculatedusingthemethoddescribed galaxies first investigated by us, there are only 6 HC3N detec-
inAppendixA,andareshowninTable7.Asalreadymentioned, tionsin the 19 galaxies.This sample of galaxieswas neverthe-
theHC3N/HCNandHC3N/HNClineratiosofsomeofthemost less made to find a high number of HC3N-luminous galaxies.
nearby galaxies will be somewhat overestimated due to their Thiscouldmeanoneofthreethings:
sourcesizebeinglargerthanthebeamsizeofthetelescope.See
– Oursearch-criteriaforHC N-luminousgalaxiesarenotap-
AppendixA.1 for a discussion on this subject. The galaxies in 3
our survey that do not fulfil the criterion θ . θ are IC 342, propriate.
s mb
M82,Maffei2,andNGC253,andwethereforeexpectthereal – ThelimitforHC3N-luminousgalaxiesissettoohightoin-
HC N/HCN and HC N/HNC ratios to be somewhat lower for cludeall“interesting”objects.
3 3
– HC N-luminous galaxies are very rare, even among active
thesegalaxies. 3
galaxies.
In Table 7, preference has been given to HC N 10-9 lines
3
before other HC N lines. Only if no 10-9 line is available, an-
3
SeveralpossiblecorrelationsbetweenahighHC Nintensityand
3
2 http://www.chalmers.se/rss/oso-en/observations/data-reduction- otherpropertiesofthegalaxieshavebeenexamined.Thiswillbe
software discussedbelow.
4
J.E.Lindbergetal.:ASurveyofHC NinExtragalacticSources
3
Table3.Listofinvestigatedobjectsandsomeoftheirproperties.Forthegalaxieswithnewobservationsreported,thegivenpositions
arethoseusedforourobservations.ForobjectsnotobservedbyusthepositioninNED(2009)isgiven.
Galaxy R.A. Decl. Typea czb Dc logL d θ e
FIR HCN
(J2000.0) (J2000.0) [kms−1] [Mpc] [L⊙] [′′]
Arp220 153457.1 +233011.3 ULIRG,Obsc.,SB? 5450 78.1 12.15 21
Circinus 141034.3 –645212.5 AGN,cp? 434f 3.13 ... 21(CO3-2)2
IC342 034648.5 +680546 SB 31 4.00 10.01 203
IC694g 112833.6 +583346.0 SB 3159 58.2 11.74h 54
IC860 131503.5 +243708.0 Obsc. 3887 59.1i 11.14 *
I17208j 172321.9 –001700.9 ULIRG,Obsc.,SB? 12852 178 12.35 0.675
M82 095552.7 +694046 SB 187 5.68 10.61 >301
Maffei2 024155.1 +593615.0 SB -17f 3.34 ... 20×76
NGC34 001106.5 –120626.6 SB 5931 79.8 11.34 *
NGC253 004733.1 –251718 SB 261 3.22 10.29 18×87
NGC613 013636.7 –290950.4 cp 1475 18.6 10.22 40(CO1-0)8
NGC1056 024248.3 +283427.1 AGN 1545 22.7 9.79 -
NGC1068 024240.7 –000047.0 cp 1005 15.3 10.89 4.51
NGC1365 033336.4 –360826.1 cp 1636 19.9 10.86 34(CO2-1)9
NGC1377 033639.1 –205408.0 Obsc.,AGN? 1792 22.5 9.95 -
NGC1614 043624.2 –082840.3 SB 4746 63.4 11.43 12(CO2-1)10
NGC1808 050742.3 –373047 SB,cp? 1000 11.2 10.55 18(CO2-1)10
NGC2146 061837.8 +782122.9 SB 885 16.9 10.93 201
NGC2623 083824.1 +254517.2 SB 5538 76.9 11.48 1.811
NGC3079 100157.81 +554047.1 SB?AGN? 1142 19.7 10.65 13×5(CO1-0)12
NGC3256 102751.3 –435414 SB 2781 36.5 11.43 9(CO2-1)10
NGC3690g 112831.0 +583340.0 SB 3159 46.9 11.74h 1.565
NGC4418 122654.8k –005242.0k Obsc.,AGN? 2104 32.6 11.00 5(CO1-0)13
NGC4945 130527.0 –492804.5 SB,cp? 560 4.85 10.41 15(CO3-2)2
NGC5135 132544.0 –295002.2 cp 4114 56.0 11.06 15×5(CO1-0)14
NGC6946 203452.3 +600914.0 SB 53 5.64 10.01 101
NGC7130 214819.5 –345705 cp 4824 65.4 11.23 10(CO1-0)10
UGC5101 093551.6 +612111.7 LIRG,cp 11785 165 11.87 3.50(CO1-0)5
References.(1)TableinKripsetal.(2008);(2)TableinCurranetal.(2001a);(3)HCNmapinMeier&Turner(2005);(4)HCNmapinAaltoetal.
(1997);(5)TableinGracia´-Carpioetal.(2008);(6)HCNmapinNguyen-Rieuetal.(1994);(7)HCNmapinKnudsenetal.(2007);(8)CO1-0
map in Bajajaetal. (1995); (9) CO 2-1 source size inCurranetal. (2001b); (10) Table inAaltoetal. (1995); (11) Table in Bryant&Scoville
(1999);(12)CO1-0mapinKodaetal.(2002);(13)CO1-0mapinDaleetal.(2005);(14)CO1-0mapinReganetal.(1999).
Notes. (a) The classificationshave been obtained by careful investigation of thenotes in NED (2009). SB = starburst, AGN = activegalactic
nucleus, cp = composite of SB and AGN, Obsc. = obscured, ULIRG = Ultra-luminous Infrared galaxy, LIRG = Luminous Infrared galaxy.
(b) Heliocentric radial velocity of source, from Sandersetal. (2003). (c) Distance to source, corrected for Virgoinfall only, from NED (2009).
(d) FarInfraredLuminosity,fromSandersetal.(2003).(e) Sourcesizes,givenforHCN1-0lineifnotspecifiedotherwise.Forgalaxieswithan
asterisk(*),novaluehasbeenfound,butD&45Mpc,allowingthepoint-likeapproximation(θ =0)withanerror.5%ifthedensemolecular
HCN
gasinthisgalaxyisnotunusuallywidelydistributed.Forgalaxiesmarkedwithadash(-),novaluehasbeenfound,andD < 45Mpc.(f) From
NED(2009).(g) IC694andNGC3690aretogetheralsoknownasthemergerArp299.(h) ThisistheFIRluminosityofIC694andNGC3690
together.(i)FromSandersetal.(2003).(j) ShortforIRAS17208-0014.(k)FortheHCN1-0datathefollowingcoordinateshavebeenused:R.A.
122654.63,Decl.–005239.6(J2000.0).
5.1.InwhichtypesofgalaxiesdowefindHCN? 5.2.Silicateabsorptionstrength
3
InSpoonetal.(2002),severalabsorptionfeaturesfromice and
If the HC N-rich and -poor galaxies are compared with the
3 silicates as well as emission from PAHs in active galaxies are
galaxyclassificationsofTable3,the mostobvioustrendisthat
discussed.InSpoonetal.(2007),anevolutionaryplotforactive
mostofthe HC N-poorgalaxiesarestarbursts, withtheexcep-
3 galaxiesisproduced,showingtwodistinctregionsinaplotover
tionfortheLIRGUGC5101.Itisdifficultto seeanytrendfor
theequivalentwidthofthePAH6.2µmemissionlineversusthe
theAGNsduetothelownumberofsuchobjectsinthesample.
strengthofthesilicate9.7µmabsorptionband.Starburstgalax-
When removing the nearby galaxies from the HC N-rich ies tend to have a high PAH equivalent width, Seyfert galax-
3
category due to their overestimated HC N/HCN-ratios men- ies have low PAH equivalentwidth and low silicate absorption
3
tionedearlier,thecommondenominatoroftheremaininggalax- strength, while ULIRGs have high silicate absorption strength
ies seems to be that their source of activity is unknownor dis- andoftenalsolowPAHequivalentwidth.
puted–theyarelabelledas“obscured”orULIRGs.Thus,HC N By private communication with H. W. W. Spoon, the nu-
3
mightthriveindeeplyobscured,shieldedregions,whereitcan- merical values for silicate absorption strength in all the galax-
notbedestroyedbyradiation.Instarbursts,itmightbedestroyed iesinhissamplewereobtained.Mostgalaxiesinoursampleare
by the strong UV field – or not even created, as C H on the alsoincludedinhissample.Whencomparingthesevaluestoour
2 2
grainswillphoto-dissociateintoC H(seeSection2.1). HC N/HCNratios,atentativepatternseemstoappear.
2 3
5
J.E.Lindbergetal.:ASurveyofHC NinExtragalacticSources
3
Table4.DatafromHC Nobservations.
3
Galaxy Line I(HC N)a S ∆v ∆v Telescope η θ T scalea References
3 ν mb mb
[Kkms−1] [Jykms−1] [kms−1] [′′]
Arp220 10-9 2.02±0.15 45±3 340 OSO20m 0.59 44 T∗ (1)
A
Arp220 10-9b 0.4±0.15 11±4 350 SEST15m 0.75 55 T∗ (2)
Arp220 12-11 0.96±0.1 23±2 170 OSO20m 0.52 36 TA∗ (1)
A
Circinus 10-9b 1.01±0.1 27.6±2.7 290 SEST15m 0.75 55 T∗ (1)
IC342 10-9 2.6±0.7 14±4 52 IRAM30m 0.8 25 TA∗ (3)
IC694 12-11 <0.30 <7.4 ... OSO20m 0.52 36 TR∗ (1)
IC860 28-27 0.54±0.07 3.9±0.5 175 IRAM30m 0.59 9 TA∗ (1)
I17208 10-9 0.33±0.03 2.2±0.2 330 IRAM30m 0.80 28 TA∗ (1)
M82 12-11 5.6±0.6 30±3 155 IRAM30m 0.80 25 TA∗ (3)
Maffei2 12-11 1.42±0.05 10.9±0.4 200 IRAM30m 0.73 24 TR∗ (1)
A
NGC34 10-9b <0.45 <12 ... SEST15m 0.75 55 T∗ (1)
A
NGC253 9-8 5.8±0.6 27±3 63c IRAM30m ... 29 T (4)
mb
NGC253 10-9 5.8±0.6 27±3 63c IRAM30m ... 26 T (4)
mb
NGC253 12-11 4.4±0.7 19±3 63c IRAM30m ... 21 T (4)
mb
NGC253 15-14 4.4±0.3 24±2 77,85 IRAM30m ... 19 T (5)
mb
NGC253 15-14 3.6±0.6 16±3 63c IRAM30m ... 17 T (4)
mb
NGC253 16-15 3.8 19 77c,85c IRAM30m ... 17 T (5)
mb
NGC253 17-16 3.0±0.2 15±1 72c IRAM30m ... 16 T (5)
mb
NGC253 17-16 3.4±0.4 15±2 63c IRAM30m ... 15 T (4)
mb
NGC253 18-17 2.2±0.5 11±2 73 IRAM30m ... 15 T (5)
mb
NGC253 19-18 4.6±0.6 22±3 74 IRAM30m ... 14 T (5)
mb
NGC253 26-25 3.2±0.7 21±5 63c IRAM30m ... 12 T (4)
mb
NGC613 10-9b <0.26 <7.0 ... SEST15m 0.75 55 T∗ (1)
NGC1056 16-15 <0.17 <1.3 ... IRAM30m 0.67 17 TA∗ (1)
A
NGC1068 10-9b 0.39±0.05 11±1 100 SEST15m 0.75 55 T∗ (1)d
A
NGC1365 10-9b <0.61 <417 ... SEST15m 0.75 55 T∗ (1)d
NGC1377 16-15 <0.28 <2.1 ... IRAM30m 0.67 17 TA∗ (1)
NGC1377 25-24 <0.26 <1.9 ... IRAM30m 0.63 10.5 TA∗ (1)
A
NGC1614 10-9b <0.39 <11 ... SEST15m 0.75 55 T∗ (1)
A
NGC1808 10-9b 0.2±0.1 6±3 250 SEST15m 0.75 57 T∗ (2)
NGC2146 10-9 <0.38 <7.7 ... OSO20m 0.59 42 TA∗ (1)
NGC2146 12-11 <0.38 <9.2 ... OSO20m 0.52 36 TA∗ (1)
NGC2623 12-11 <0.50 <12 ... OSO20m 0.52 36 TA∗ (1)
NGC3079 10-9 0.60±0.05 4.0±0.3 500 IRAM30m 0.80 28 TA∗ (1)
NGC3079 16-15 <0.54 <4 ... IRAM30m 0.80 17 TA∗ (1)
NGC3079 25-24 <0.40 <3 ... IRAM30m 0.80 10.5 TA∗ (1)
A
NGC3256 10-9b <0.12 <3.3 ... SEST15m 0.75 55 T∗ (2)
NGC3690 12-11 <0.31 <7.5 ... OSO20m 0.52 36 TA∗ (1)
NGC4418e 10-9 0.8±0.08 5±0.5 122 IRAM30m 0.77 27 TA∗ (6)
A
NGC4418e 16-15f 1.7±0.08 12±0.6 130 IRAM30m 0.70 17 T∗ (6)
NGC4418e 25-24 1.6±0.2 15±2 140 IRAM30m 0.53 11 TA∗ (6)
A
NGC4945 9-8 2.16±0.50 47±11 230 SEST15m 0.78 63 T (7)
mb
NGC4945 10-9 1.99±0.21 41±4 290 SEST15m 0.75 55 T (7)
mb
NGC4945 11-10 2.92±0.35 65±8 340 SEST15m 0.73 52 T (7)
mb
NGC4945 12-11 4.18±0.38 98±9 340 SEST15m 0.71 49 T (7)
mb
NGC4945 15-14 2.13±0.29 52±7 250 SEST15m 0.65 40 T (7)
mb
NGC4945 16-15g 5.02±0.19 120±5 330 SEST15m 0.63 37 T (7)
mb
NGC4945 17-16 2.26±0.55 48±12 280 SEST15m 0.61 33 T (7)
mb
NGC4945 24-23 <0.60 <12 ... SEST15m 0.48 23 T (7)
mb
NGC4945 25-24 <0.60 <12 ... SEST15m 0.46 22 T (7)
mb
NGC5135 10-9b <0.13 <3.7 ... SEST15m 0.75 55 T∗ (1)
NGC6946 12-11 <0.32 <2.5 ... IRAM30m 0.73 24 TA∗ (1)
A
NGC7130 10-9b <0.10 <2.7 ... SEST15m 0.75 55 T∗ (2)
UGC5101 10-9 <0.12 <0.76 ... IRAM30m 0.80 28 TA∗ (1)
A
References.(1)Thiswork;(2)Aaltoetal.(2002);(3)Henkeletal.(1988);(4)Mauersbergeretal.(1990);(5)Mart´ınetal.(2006);(6)Aaltoetal.
(2007);(7)Wangetal.(2004).
Notes. (a)ThetemperaturescaleoftheintegratedintensitiesaregivenintheT scalecolumn.Upperlimitsare2σcalculatedfromthermsofthe
noisesurroundingthelineforour data.Forour data,errorsaregivenin1σandcalculatedfromtherms.Sincemanyarticleslackinformation
aboutsizesoferrorsandmethodsusedwhencalculatingtheerrors,mosterrorsaregivenasprintedintherespectivearticle.However,ifthesizeof
theerrorisclearlywritten,ithasbeenrecalculatedto1σ.(b)MeasuredinHNC1-0spectrum.(c)FixedwhenfittingGaussian.(d)Thesevaluesare
calculatedfromHNC1-0spectraalreadypublishedinPe´rez-Beaupuitsetal.(2007).(e)SeeCostagliola&Aalto(2010a)foranextensivesurvey
ofHC Ninthisgalaxy,withdatanotincludedinthistable.(f) Contaminatedbypara-H CO,estimatedto20%.Thegivenvalueisonlyforthe
3 2
HC Ncomponent.(g)Contaminatedbypara-H CO.
3 2
6
J.E.Lindbergetal.:ASurveyofHC NinExtragalacticSources
3
Table5.DatafromHCN1-0observations
Galaxy I(HCN)1-0a S ∆v ∆v Telescope η θ T scalea References
ν mb mb
[Kkms−1] [Jykms−1] [kms−1] [′′]
Arp220 9.7±0.4 57±2 530 IRAM30m 0.82 29.5 T (1)
mb
Circinus 5.2±0.8 110±20 300 SEST15m 0.75 57 T (2)
mb
IC342 15.5±0.9 36±2 ... NRO45m 0.54 19 T (3)
mb
IC694 1.29±0.09 8.4±0.6 ... IRAM30m 0.82 28 T∗ (4)
A
IC860 ... ... ... ... ... N/Ab
I17208 2.19±0.16 14±1 ... IRAM30m 0.82 28 T∗ (4)
I17208 0.91±0.19 34±7 ... NRAO12m 0.89 72 TA∗ (5)
R
M82 29±0.2 170±1 130 IRAM30m 0.82 29.5 T (1)
mb
Maffei2 13.8±0.9 32±2 ... NRO45m 0.54 19 T (3)
mb
NGC34 1.6±0.2 33±4 600-700 SEST15m 0.75 57 T (6)
mb
NGC253 40.8 175 150 NRO45m 0.45 23 T∗ (7)
NGC613 0.53±0.08 15±2 130 SEST15m 0.75 57 TA∗ (8)
A
NGC1056 ... ... ... ... ... N/Ab
NGC1068 24.5±0.9 145±5 220 IRAM30m 0.82 29.5 T (1)
mb
NGC1365 6.0±0.1 125±2 300-400 SEST15m 0.75 57 T (6)
mb
NGC1377 0.47±0.1 3.0±0.6 140 IRAM30m 0.80 28 T∗ (8)
NGC1614 1.5±0.22 40±6 300 FCRAO14m 0.60 50 TA∗ (5)
NGC1808 4 110 SEST15m 0.74 56 TA∗ (9)
A
NGC2146 5±0.1 30±1 290 IRAM30m 0.82 29.5 T (1)
mb
NGC2623 ... ... ... ... ... N/Ab
NGC3079 5.7±0.8 29±4 420 IRAM30m 0.80 28 T (10)
mb
NGC3079 2.6±0.42 97±16 365 NRAO12m 0.89 72 T∗ (5)
R
NGC3256 2.3±0.2 48±4 165 SEST15m 0.77 57 T (11)
mb
NGC3690 2.04±0.11 13.2±0.7 300 IRAM30m 0.82 28 T∗ (4)
NGC4418 1.96±0.04 12.4±0.3 170 IRAM30m 0.80 28 TA∗ (8)
A
NGC4945 22.4±0.4 436±8 305 SEST15m 0.75 55 T (12)
mb
NGC5135 0.65±0.07 14±1.5 50-60 SEST15m 0.75 57 T (6)
mb
NGC6946 8.7±0.9 20±2 ... NRO45m 0.54 19 T (3)
mb
NGC7130 0.7±0.1 15±2 100 SEST15m 0.75 57 T (6)
mb
UGC5101 1.40±0.14 9.1±0.9 500 IRAM30m 0.82 28 T∗ (4)
A
References.(1)Kripsetal.(2008);(2)Curranetal.(2001a);(3)Soraietal.(2002);(4)Gracia´-Carpioetal.(2008);(5)Gao&Solomon(2004);
(6) Curranetal. (2000); (7) Nguyen-Q-Rieuetal. (1989); (8) This work; (9) Aaltoetal. (1994); (10) Pe´rez-Beaupuitsetal. (2007); (11)
Casolietal.(1992);(12)Wangetal.(2004).
Notes. (a)ThetemperaturescaleoftheintegratedintensitiesaregivenintheT scalecolumn.Upperlimitsare2σcalculatedfromthermsofthe
noisesurroundingthelineforour data.Forour data,errorsaregivenin1σandcalculatedfromtherms.Sincemanyarticleslackinformation
aboutsizesoferrorsandmethodsusedwhencalculatingtheerrors,mosterrorsaregivenasprintedintherespectivearticle.However,ifthesize
oftheerrorisclearlywritten,ithasbeenrecalculatedto1σ.(b)Nosingle-dishHCNdatawerefoundintheliteraturefortheseobjects.
In Figure 1, the relation between the HC N/HCN ratio and moleculargasismorelikelytoharbourastrongmegamaserthan
3
the silicate absorption strength is plotted. We note that two of onewithlittle moleculargas–thedatainDarling(2007)show
the three nearby galaxies in the sample, M82 and IC 342, are a strong correlation between OH megamaser strength and CO
showingtoohighHC N/HCNratiostofitintothepatternofthe and HCN luminosity. Thus, it is more reasonable to normalise
3
figure, which was expected (see Appendix A.1). A correlation theOHmegamaserluminositywithsomekindofluminosityfor
seemspossiblewhenexcludingthenearbyM82,NGC253,and themoleculargasinthegalaxy.TheCO1-0luminosityhasbeen
IC 342 (correlationcoefficient r = −0.49). One explanation to chosenforthis,sinceCOistheprimarytracerofmoleculargas.
the correlation might be that HC N is formed in regions with For IC 860, no CO 1-0 luminosity value has been found
3
silicates, where the silicates protect the HC N from radiation. in the literature, but from an observation performed with the
3
Thus,HC Nsurvivesbetterinregionsheavilyobscuredbysili- IRAM 30 m telescope by F. Costagliola the luminosity could
3
cates. be calculated from the intensity with the method described in
Solomonetal.(1997).Theusedequationis:
5.3.Megamasers LCO =23.5Ωs∗bD2LICO(1+z)−3 (7)
where LCO is the line luminosity in K km s−1 pc2, Ωs∗b is the
Neglectingthemostnearbygalaxies,wheretheHC N/HCNra- solid angleofthe sourceconvolvedwith thetelescope beamin
3
tios probablyare somewhat overestimated,all HC N-luminous arcsec2, D is the luminosity distance in Mpc, I is the main
3 L CO
galaxies have OH mega- or kilomasers (Darling&Giovanelli beamintensityofthelineinKkms−1,andzistheredshiftofthe
2006).AfewoftheHC N-poorgalaxiesalsohaveOHmega-or source.Theintensity9.83Kkms−1,thebeamwidth22′′,andthe
3
kilomaseractivity(NGC253,NGC1068,andUGC5101).The distance59.1Mpcgivesaluminosityof3.07·108Kkms−1pc2,
conceptofdefiningmegamaserstrengthonlyfromitsluminos- assumingthatthesourcesizeismuchsmallerthanthebeamsize,
ityishoweversomewhatmisleading.Firstly,agalaxywithmuch whichisvalidforIC860.
7
J.E.Lindbergetal.:ASurveyofHC NinExtragalacticSources
3
Table6.DatafromHNC1-0observations
Galaxy I(HNC)1-0a S ∆v ∆v Telescope η θ T scalea References
ν mb mb
[Kkms−1] [Jykms−1] [kms−1] [′′]
Arp220 0.95±0.2 26±6 ... SEST15m 0.75 55 T∗ (1)
Circinus 1.99±0.1 55±3 280 SEST15m 0.75 55 TA∗ (2)
IC342 9.2±0.7 48±4 47 IRAM30m 0.8 25 TA∗ (3)
IC694 0.75±0.2 15±4 300-400 OSO20m 0.59 42 TR∗ (1)
IC860 0.70±0.04 4.6±0.3 230 IRAM30m 0.80 28 TA∗ (2)
I17208 1.12±0.06 7.4±0.4 350 IRAM30m 0.80 28 TA∗ (2)
A
M82 7.3±0.6 31±3 129 IRAM30m 0.64 25 T (4)
mb
Maffei2 7.3±0.8 31±3 91,50 IRAM30m 0.64 25 T (4)
mb
NGC34 <0.44 <12 ... SEST15m 0.75 55 T∗ (2)
A
NGC253 50.0±2.8 210±12 72,136 IRAM30m 0.64 25 T (4)
mb
NGC613 <0.24 <6.5 ... SEST15m 0.75 55 T∗ (2)
NGC1056 <0.17 <1.1 ... IRAM30m 0.80 28 TA∗ (2)
A
NGC1068 3.2±0.5 65±10 260 SEST15m 0.75 55 T (5)
mb
NGC1068 11.4±0.7 48±3 232 IRAM30m 0.64 25 T (4)
mb
NGC1365 4.7±0.6 96±12 150 SEST15m 0.75 55 T (5)
mb
NGC1377 <0.15 <0.99 ... IRAM30m 0.80 28 T∗ (2)
NGC1614 <0.38 <10 ... SEST15m 0.75 55 TA∗ (2)
NGC1808 1.2±0.1 33±3 300 SEST15m 0.75 55 TA∗ (1)
A
NGC2146 1.6±0.3 6.7±1.3 237 IRAM30m 0.64 25 T (4)
mb
NGC2623 0.6±0.15 12±3 500-600 OSO20m 0.59 42 T∗ (1)
A
NGC3079 2.9±0.5 15±3 380 IRAM30m 0.80 28 T (5)
mb
NGC3079 6.9±1.0 29±4 545 IRAM30m 0.64 25 T (4)
mb
NGC3256 0.6±0.05 16±1 250 SEST15m 0.75 55 T∗ (1)
A
NGC3690 ... ... ... ... ... N/Ab
NGC4418 1.24±0.12 7.9±0.8 156 IRAM30m 0.77 27 T∗ (6)
NGC4945 8.6±0.2 230±5 290 SEST15m 0.75 55 TA∗ (2)
NGC5135 <0.13 <3.5 ... SEST15m 0.75 55 TA∗ (2)
A
NGC6946 4.0±0.3 17±1 138 IRAM30m 0.64 25 T (4)
mb
NGC7130 0.4±0.05 11±1 ... SEST15m 0.75 55 T∗ (1)
UGC5101 1.24±0.1 8.2±0.7 500 IRAM30m 0.80 28 TA∗ (2)
A
References.(1)Aaltoetal.(2002);(2)Thiswork;(3)Henkeletal.(1988);(4)Hu¨ttemeisteretal.(1995);(5)Pe´rez-Beaupuitsetal.(2007);(6)
Aaltoetal.(2007).
Notes. (a)ThetemperaturescaleoftheintegratedintensitiesaregivenintheT scalecolumn.Upperlimitsare2σcalculatedfromthermsofthe
noisesurroundingthelineforour data.Forour data,errorsaregivenin1σandcalculatedfromtherms.Sincemanyarticleslackinformation
aboutsizesoferrorsandmethodsusedwhencalculatingtheerrors,mosterrorsaregivenasprintedintherespectivearticle.However,ifthesize
oftheerrorisclearlywritten,ithasbeenrecalculatedto1σ.(b)NoHNCdatawerefoundintheliteratureforthisobject.
In Table 8, the OH megamaser luminosity is compared UV radiation by the warm dust which is needed to power the
with the CO luminosity. The data in this table is also shown megamaser(Darling&Giovanelli2006).Anotherpossibilityis
in the form of a histogram in Figure 2. The non-detections that the HC N is pumped by the IR field caused by the warm
3
of OH megamasers and OH absorbers reported in Darling dust,whichalsopumpstheOHmegamaser.
(2007) are also listed, most of them being either HC N-poor
3
or withoutany HC N detection(the exceptionsare IC 342and
3
M82,whoseHC N/HCNratiosprobablyareoverestimated,see 5.4.IRfluxdensityratios
3
AppendixA.1).
In several of the HC N rich galaxies, rotational-vibrational
3
It can be seen that the OH megamaser luminosities nor- HC N lines have been detected (Costagliola&Aalto 2010a;
3
malisedwiththegalacticCOluminositiesaremuchhigherinthe Mart´ınetal. 2010; Costagliolaetal. 2010b). Thissuggests that
HC3N-luminousgalaxies than in the HC3N-poor galaxies with IRpumpingofthe emission is presentin these galaxies,which
anOHmegamaser,especiallywhenignoringthenearbygalaxies in turn indicates a warmer spectral energy distribution (SED)
M82andIC342.TheaverageoftheOH/COluminosityratiosis in these sources. We have thus compared the HC N/HCN ra-
3
morethan10timeshigherintheHC3N-luminousgalaxiesthan tios with the IRAS 60 µm/100 µm flux density ratios, but no
in the HC3N-poor galaxies. The non-detections(where HC3N- linear correlation could be found. However, when plotting the
poorgalaxiesareoverrepresented)arenotincludedintheseav- data as a histogram (Figure 3), we see a trend towards HC N-
3
erages. UGC 5101 is the only HC3N-poor galaxy with an OH luminous galaxies having higher IRAS 60 µm/100 µm ratios,
megamaserthatisstrongcomparedtotheamountofmolecular correspondingtowarmerSEDs.NGC4418,theobjectwiththe
gasin the galaxy.Also withoutthe CO normalisationthe trend highestHC N/HCN ratio in the sample is also the galaxy with
3
canbeseenclearly. thehighestIRAS60µm/100µmratio.
A possible explanation of the HC N correlating with OH It should however be mentioned that the global
3
megamasers is that the HC N is protected against destructive 60 µm/100 µm flux density ratios might not be completely
3
8
J.E.Lindbergetal.:ASurveyofHC NinExtragalacticSources
3
0.45 8
NGC 7130 HC3N-faint galaxy
0.40 NGC 4418 7 NOetharebr yH HCC3N3N-l-ulmuminionuosu sg aglaalxayxy
0.35 NGC 5135 Galaxy with a high HC3N/HCN upper limit
0)0.30 xies6
log(HCN10/HCN1 3(cid:1)(cid:1) 000...212055 IICC 334422 MM8822 NGC I310772908 Arp 220 ulatednumberofgala 345 NNNGGGCCC 124619144465
m IC 694
0.10 NGC 1068 NNGGCC 225533 NGC 4945 Accu2 NGC 3079 NGC+ 3690 I17208
0.05 NGC 1365 NGC 1068 UGC 5101 Arp 220
1
0.00
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5
(cid:0) (cid:0) (cid:0)Silicate(cid:0) absorpti(cid:0)on strengt(cid:0)h (cid:0) (cid:0) (cid:0) IC 342 M82 NGC 253 IC 860 NGC 4418
0
no OH det. 3 2 1 0 1
Fig.1. Tentative correlationbetween HC3N/HCN ratio and sil- (cid:2) logLL(cid:2)COOH [108KkLm(cid:3)s(cid:4)1p(cid:2)c2 ]
icate absorption strength. The HC N/HCN values for IC 342,
3
NGC 253, and M82 are probably overestimated (indicated by Fig.2. Histogram over number of galaxies in normalised OH
thinnererrorbars,seeAppendixA.1).Theobjectsforwhichthe megamaserluminosity.Light-greycorrespondstogalaxiescon-
HC3N 12-11 line hasbeen used instead of the HC3N 10-9 line sideredHC3N-poor,medium-graycorrespondstonearbyHC3N-
when calculating the HC3N/HCN ratio are indicated by a star. luminousgalaxieswhoseHC3N/HCN ratiosprobablyareover-
Thesilicateabsorptionstrengthisinamagnitudescale:Ahigher estimated(seeAppendixA.1),dark-greycorrespondstoallother
negativenumbermeansstrongersilicateabsorption. HC3N-luminous galaxies, and white corresponds to galaxies
wheretheupperlimitontheHC N/HCNratioistoohightotell
3
whetheritisHC N-luminousornot.ThegalaxiesforwhichOH
3
relevant,sincetheymaytellmoreaboutthetemperatureonthe maser emissionis non-detectedin Darling&Giovanelli(2002)
extendeddustdistribution(and/orforegrounddust)thanthedust and/orDarling(2007)arestackedintheleftmostcolumnofthe
temperatureinthenucleus,wherethepumpinglikelyoccurs. histogram.
5.5.Ciiflux
8
ManyoftheobjectswithhighHC3N/HCNratiosalsohavelow HC3N-faint galaxy NGC 3690
Cii/FIRfluxratios.ThiscanbeexplainedbythattheC+ionsare 7 Nearby HC3N-luminous galaxy
WabelehtaovedessetarrocyheHdCi3nNthtehrloituegrahturreeacfotironavsa5ilaabnlde6CiiniflSuexcetisoonf2t.h2e. xies6 OHitghhe rH HCC33NN/-HluCmNi nuopupse gr alliamxiyt NGC 2623
objects, which were found for more than half of the objects in ala NGC 7130 NGC 1614
oursample.Withthe limitedamountofdata,nolinearcorrela- ofg 5
tioncouldbeestablished.Instead,wedisplaythedataintheform ber NGC 613 NGC 2146 NGC 3256
ofahistograminFigure4,whereweclearlyseethatamajority num4
swoefeethtehxeaptHetchCtee3dNne-taorircbbhyeggaaanllaaoxxviieeerss-feaosrrteiwmvhaeitrciyohnpthodeourheiitgnohCbHeiCiam3flNue/xHff.eCWcNtesr(aasltesiooe Accumulated 23 NNGGCC 34097495 NNGGCC 15015365 NNGGCC 1 205638 AICrp 3 24220 NGC 1377
AppendixA.1) allbelongto the Cii rich partof the histogram.
NGC 6946 UGC 5101 NGC 1808 I17208 NGC 34
Also in this case, NGC 4418 is the most extreme galaxy, with 1
thelowestupperlimitontheCii/FIRfluxratio. Maffei 2 NGC 1365 Circinus IC 860 M82 NGC 4418
0
0.4 0.3 0.2 0.1 0 0.1 0.2
(cid:5) (cid:5) (cid:5) log I(cid:5)RAS60(cid:6)m
5.6.HNC/HCN IRAS100m
(cid:6)
The HNC/HCN 1-0 ratio is an indicator of the physical and Fig.3. Histogram over number of galaxies in IRAS
chemicalconditionsin the dense molecular gas. Overluminous 60 µm/100 µm flux density ratio. Light-grey corresponds
HNC is a sign of XDR chemistry (Aalto 2008), while a low to galaxies considered HC3N-poor, medium-gray corresponds
HNC/HCN ratio indicates shocks (Schilkeetal. 1992). As is to nearby HC3N-luminous galaxies whose HC3N/HCN ra-
seeninFigure5,wefindacorrelationbetweentheHC N/HCN tios probably are overestimated (see Appendix A.1), dark-grey
3
andHNC/HCN 1-0lineratios.TheHC3N/HCN ratiosofM82, correspondstoallotherHC3N-luminousgalaxies,andwhitecor-
Maffei2,IC342,andNGC253areprobablyoverestimated(see responds to galaxies where the upper limit on the HC3N/HCN
AppendixA.1),andthecorrelationcoefficientisr = 0.65when ratio is too high to tell whether it is HC3N-luminous or
thesefourobjectsareexcluded.Thiscorrelationwillbefurther not. IRAS 60 µm/100 µm flux density values are all from
discussedinCostagliolaetal.(2010b). Sandersetal. (2003), except for Circinus and Maffei 2, which
Attemptswerealsomadetryingtofindacorrelationbetween arefromBeichmanetal.(1988).
theHC NratiosandratiosofhigherHNCandHCNtransitions,
3
e.g. the HNC/HCN 3-2 ratio. However, too few HNC 3-2 and
9
J.E.Lindbergetal.:ASurveyofHC NinExtragalacticSources
3
10 0.2
(cid:8)
HC3N-faint galaxy NGC 3690
9 Nearby HC3N-luminous galaxy
Other HC3N-luminous galaxy NGC 1614 (cid:8)0.4 NGC 4418
8 High HC3N/HCN upper limit
Accumulatednumberofgalaxies 34567 NGC 4418 NNNNNMGGGGGaCCCCCf f61 14e29039i5 4664236855 NNGGCC 32215466 log(HCN10/HCN10) 3(cid:10)(cid:10) (cid:8)(cid:8)(cid:8)(cid:8)0101....6082 M82 NGC 1808 NAGrp CC 2i1r 2 c0NN0i 6nGIG18uCC7s 2 3400987495 NGC 253 Maffei 2 IC 342
2
1.4
IC 860 M82 NGC 3079 (cid:8)
1
Arp 220 I17208 Circinus IC 342 1.6
0 (cid:7)3.6 lo(cid:7)g3F.C2II (cid:7)2.8 (cid:7)2.4 (cid:8) (cid:8)0.7 (cid:8)0.6 (cid:8)0.5 (cid:8)lo0g.(4HNC 1(cid:8)(cid:9)00./3HCN 1(cid:8)(cid:9)00.)2 (cid:8)0.1 0.0 0.1
FFIR Fig.5. log-log plot of the HC N/HCN line ratio versus the
Fig.4. Histogram over number of galaxies in Cii flux nor- HNC/HCNlineratio.TheHC N3/HCNvaluesforIC342,Maffei
3
malised by FIR flux. Light-grey corresponds to galaxies con- 2,M82,andNGC253areprobablyoverestimated(indicatedby
sideredHC3N-poor,medium-graycorrespondstonearbyHC3N- thin error bars, see Appendix A.1). The objects for which the
luminousgalaxieswhoseHC3N/HCN ratiosprobablyareover- HC3N 12-11 line hasbeen used instead of the HC3N 10-9 line
estimated (see Appendix A.1), dark-grey corresponds to all whencalculatingtheHC N/HCNratioareindicatedbyastar.
3
otherHC N-luminousgalaxies,andwhitecorrespondstogalax-
3
ies where the upper limit on the HC N/HCN ratio is too
3
high to tell whether it is HC N-luminous or not. FIR fluxes AnewlinesurveywasmadewiththenewEMIRreceiveratthe
3
were calculated from IRAS 60µm and 100µm fluxes with the IRAM30mtelescopeinJune2009,andtheresultsfromthissur-
method described in Bizyaev (2001). The references for the veywillbe publishedin Costagliolaetal. (2010b). Thesample
used IRAS fluxes are given in the caption of Figure 3. Cii inthissurveyhasbeenchosentogetmoreHC Ndataonsources
3
flux values are from Negishietal. (2001), except for IC 342 in Spoon’ssample to test the possible correlationsbetween the
and NGC 3079 (Staceyetal. 1991); IC 860 and NGC 4418 HC N/HCN ratioandPAH equivalentwidth and/orsilicate ab-
3
(Malhotraetal.2001);Arp220(Luhmanetal.2003);andNGC sorption strength. The bandwidth of the EMIR receiver allows
1614(Brauheretal.2009). for severalspectrallines being observedin the same spectrum,
andthusmanydifferentmolecularspeciescanbeobservedatthe
sametime.
HCN 3-2spectraforthegalaxiesinthesampleareavailablein Itmightalsobeinterestingtosearchforotherlong-carbon-
theliterature–withonlyfivedatapointsnoconclusionscanbe chain molecules in the HC N-luminous galaxies, e.g. HC N,
3 5
drawn.ItwasalthoughnoticedthattheHCN 3-2andHNC3-2 C H,C H,C H ,andC H ,whichhaveallbeenfoundinstar-
2 4 3 2 4 2
intensities seem to be very uncertain, at least for Arp 220 and formingregionsintheGalaxy(Sakaietal.2008,2009).
NGC4418.Shortly,itseemslikethedifferentinstrumentsused Finally,wesuggestmappingofHC N,C H,HCN,andHNC
3 2
fortheobservationsaffectthemeasuredvaluetoanon-negligible inalargernumberofgalaxies,especiallyintheHC N-luminous
3
extent.ThisisfurtherdiscussedinLindberg(2009). galaxies, to compare the results with the maps of IC 342 in
Wewerenotabletoreproducetheweakcorrelationbetween Meier&Turner(2005).TheHC Nabsorptionlinesdetectedin
3
the HNC/HCN 1-0 ratio and the FIR luminosity described in az∼0.89galaxyinfrontofPKS1830-211(Henkeletal.2009)
Aaltoetal.(2002). indicate that HC N is not only present in the core of a galaxy,
3
butmayalsobepresentinthedisc.
5.7.Futureobservationaltests
6. Conclusions
WehavefoundastrongtrendbetweentheHC N/HCNratioand
3
OH megamaser activity (see Section 5.3). By studying differ- WehavepresentedthefirstsurveyofHC Nobservationsinex-
3
entexcitationlevelsofHC3NintheseandotherOHmegamaser tragalacticobjects.Themainconclusionsfromthissurveyareas
galaxies, the cause of this correlation can be investigated. If follows:
HC Nispumped,highertransitions,includingvibrationaltran-
3
sitions,shouldbefound.ThishasalreadybeendetectedinNGC 1. Bright HC N emission is rather uncommon in galaxies. It
3
4418(Costagliola&Aalto2010a). wasonlydetectedin6ofthe19galaxieswhichhadnotbeen
To better establish the importanceof HC N as an indicator investigatedbefore,eventhoughthatsamplewasselectedto
3
of activity in certain galaxies, it is importantto do further line findmanyHC N-luminousgalaxies.
3
surveysof HC N, HCN, and HNC, as well as other molecules 2. Most HC N-luminous galaxies are obscured galaxies.
3 3
who trace the properties of the molecular gas, such as HCO+ Starburstgalaxies seem to be poor in HC N. There are too
3
and C H. Such a line survey could also test our weak correla- fewAGNgalaxiesinthesampletotellifthesenormallyare
2
tion between the HC N/HCN and HNC/HCN line ratios, and richorpoorinHC N.
3 3
perhapsfinding other correlationswhich would enable a better 3. WeakcorrelationscanbeseenbetweentheHC N/HCNratio
3
understandingofthechemistryinobscuredandactivegalaxies. andsilicate9.7µmabsorptionstrength.
10