Astronomy&Astrophysicsmanuscriptno.ngc6334˙3mm (cid:13)c ESO2008 February2,2008 ATCA 3mm observations of NGC6334I and I(N): dense cores,  outflows and an UCH region H.Beuther1,A.J.Walsh2,S.Thorwirth3,Q.Zhang4,T.R.Hunter5,S.T.Megeath6 andK.M.Menten3 8 0 0 1 Max-Planck-InstituteforAstronomy,Ko¨nigstuhl17,69117Heidelberg,Germany 2 e-mail:[email protected] 2 CentreforAstronomy,JamesCookUniversity,Townsville,QLD4811Australia n e-mail:[email protected] a J 3 Max-Planck-InstituteforRadioastronomy,AufdemHu¨gel69,53121Bonn,Germany e-mail:[email protected], [email protected] 1 4 Harvard-SmithsonianCenterforAstrophysics,60GardenStreet,Cambridge,MA02138,USA 1 e-mail:[email protected] ] 5 NRAO,520EdgemontRd,Charlottesville,VA22903 h e-mail:[email protected] p 6 RitterObservatory,DepartmentofPhysicsandAstronomy,UniversityofToledo,Toledo,OH43606-3390,USA - o e-mail:[email protected] r t s a ABSTRACT [ 1 Aims.Investigationofthedensegas,theoutflowsandthecontinuumemissionfromthemassivetwincoresNGC6334IandI(N)athighspatial v resolution. 8 Methods. We imaged the region with the Australia Telescope Compact Array (ATCA) at 3.4mm wavelength in continuum as well as 7 CH CN(5 −4 )andHCN(1–0)spectrallineemission. 3 K K 7 Results.Whilethecontinuum emissioninNGC6334I mainlytracestheUCHregion, toward NGC6334I(N) wedetect lineemission from 1 four of thepreviously identifieddust continuum condensations that are of protostellar or pre-stellar nature. TheCH CN(5 −4 ) lines are . 3 K K 1 detectedinallK-componentsuptoenergiesof128Kabovegroundtowardtwoprotostellarcondensationsinbothregions.Wefindline-width 0 increasing with increasing K for all sources, which indicates a higher degree of internal motions of the hotter gas probed by these high 8 K-transitions. Toward the main mm and CH CN source in NGC6334I we identify a velocity gradient approximately perpendicular to the 3 0 large-scalemolecularoutflow.Thismaybeinterpretedasasignatureofanaccretiondisk,althoughotherscenarios,e.g.,anunresolveddouble : v source,couldproduceasimilarsignatureaswell.Nocomparablesignatureisfoundtowardanyoftheothersources.HCNdoesnottracethe i densegaswellinthisregionbutitisdominatedbythemolecularoutflows.WhiletheoutflowinNGC6334IexhibitsanormalHubble-lawlike X velocitystructure,thedataareconsistentwithaprecessingoutflowclosetotheplaneoftheskyforNGC6334I(N).Furthermore,weobservea r wide(∼15.4kms−1)HCNabsorptionline,muchbroaderthanthepreviouslyobservedCH OHandNH absorptionlines.Severalexplanations a 3 3 forthedifferencearediscussed. Keywords.techniques:interferometric—stars:earlytype—stars:formation—ISM:individual(NGC6334I&I(N))—line:profiles 1. Introduction mous with NGC6334F) and NGC6334I(N) are so interesting from a comparison point of view is that they are only sep- The massive twin cores NGC6334I and I(N) at a dis- arated by approximately 1parsec, hence they share a similar tance of 1.7kpc in the southern hemisphere (Neckel 1978; large-scalemolecularenvironment,buttheyexhibitextremely Straw&Hyland 1989) have been subjected to investigations differentcharacteristicslikelybecausetheyareatdifferentevo- formorethantwodecades.Thetworegionsarelocatedatthe lutionarystages. north-eastern end of the much larger molecular cloud/H re- gion complex NGC6334 (e.g., Rodr´ıguezetal. 1982; Gezari Both regions have been studied in much detail over the 1982; dePreeetal. 1995; Kraemer&Jackson 1999; Sandell last decades; recent summaries of the past observations can 2000;Carraletal.2002).ThereasonwhyNGC6334I(synony- be found, e.g., in Hunteretal. (2006), Beutheretal. (2007b) or Rodr´ıguezetal. (2007). Here we just outline their main Sendoffprintrequeststo:H.Beuther characteristics.NGC6334Iisaprototypicalhotmolecularcore 2 Beutheretal.:ATCA3mmobservationsofNGC6334IandI(N) right at the head of a cometary ultracompact H (UCH) Table1.Lineparameters region (dePreeetal. 1995; Kraemer&Jackson 1995). It ex- hibits rich spectral line emission (McCutcheonetal. 2000; Line Freq. E /k na u crit Thorwirthetal. 2003; Schilkeetal. 2006), a bipolar outflow [GHz] [K] [105cm−3] (Bachiller&Cernicharo 1990; Leurinietal. 2006) and H O, HCN(1–0),F=1-1 88.630 4 32 2 OH, CH OH class  and NH (3,3)/(6,6)/(8,6)/(11,9) maser HCN(1–0),F=2-1 88.632 4 32 3 3 HCN(1–0),F=0-1 88.634 4 32 emission(Moran&Rodr´ıguez1980;Forster&Caswell1989; CH CN(5 −4 ) 91.959 128 7 Gaume&Mutel1987;Brooks&Whiteoak2001;Norrisetal. 3 4 4 CH CN(5 −4 ) 91.971 78 7 1993; Caswell 1997; Walshetal. 1998; Beutheretal. 2007b; 3 3 3 CH CN(5 −4 ) 91.980 42 7 Walshetal. 2007). In contrast to that, up to very re- 3 2 2 CH CN(5 −4 ) 91.985 20 6 3 1 1 cently NGC6334I(N) was considered a typical cold core CH CN(5 −4 ) 91.987 13 6 3 0 0 since no mid-infrared and only faint near-infrared emis- For all observed lines the columns list (left to right) the sion was detected (Gezari 1982; Tapiaetal. 1996; Persietal. species/quantumnumbers,frequencies,upperlevelenergystatesE /k 2005). Furthermore, weak cm continuum and class  and (kistheBoltzmannconstant)andcriticaldensitiesn . u crit  CH3OH maser emission was reported (Carraletal. 2002; aThecriticaldensitiesncrit =A/γarecalculatedat60K(Einsteinco- Kogan&Slysh 1998; Caswell 1997; Walshetal. 1998). The efficient Aandcollisionalrateγ).ForHCN, Aandγaretakenfrom spectral line forest is considerably less dense compared to LAMBDA (Scho¨ieretal. 2005). For CH3CN, A is calculated from NGC6334I (Thorwirthetal. 2003), however, a few species A = 0.3λ−1030µ2 (λ100 inunitsof 100µmand µ = 3.9debye), and the are stronger toward NGC6334I(N) (Sollins&Megeath 2004, correspondingγ-valuesarefromPei&Zeng(1995b,a).Foralineop- tical depth τ > 1, n has to be multiplied by 1/τ (Tielens 2005). Walshetal.inprep.).Inaddition,Megeath&Tieftrunk(1999) crit report the detection of a molecular outflow in this region as well. In summary, both regions show signs of active star for- mation,however,thesouthernregionNGC6334Iappearstobe inamoreadvancedevolutionarystagethanthenorthernregion NGC6334I(N). To better characterize this intriguing pair of massive star- forming regions, we started a concerted campaign from cm cellent with approximateprecipitablewater vapor of ∼10mm to mm wavelengths with the Australia Telescope Compact and measured system temperatures between 150 and 670K. Array(ATCA),theSubmillimeterArray(SMA)andtheMopra ThephasereferencecenterswereR.A.(J2000)17h20m53s.44, single-dish telescope. The previous ATCA NH (1,1) to (6,6) Decl. (J2000)−35◦47′02′′.2 for NGC6334I and R.A. (J2000) 3 line observations revealed compact warm gas emission from 17h20m54s.63,Decl.(J2000)−35◦45′08′′.9forNGC6334I(N). bothregions(Beutheretal.2005,2007b),andtemperatureses- The velocities relative to the local standard of rest (v ) for lsr timatedtoexceed100K.WhiletowardNGC6334I(N)thelow NGC6334IandNGC6334I(N)are∼ −7.6and∼ −3.3kms−1, energy NH lines showed only extended emission, the high respectively. We observed the 3.4mm continuum emission at 3 energy lines finally revealed compact gas components. The 88.4GHzwithabandwidthandspectralresolutionof128and NH (6,6)lineprofilefromNGC6334I(N)allowedspeculation 1MHz,respectively.Thespectralrangeforthecontinuumwas 3 aboutapotentialaccretiondisk.CH OHwasstronginabsorp- checkedtobeline-freebasedonpreviousMOPRAsingle-dish 3 tiontowardthesouthernUCHregioninNGC6334I,indicative observationsofthatregion(Walshetal.,inprep.).Duringone ofexpandinggas.Inthemmcontinuumemission,Hunteretal. night, simultaneously with the continuum emission, we ob- (2006)usedtheSMAtoresolveseveralmmcontinuumsources servedtheHCN(1–0)lineat88.632GHz.Inthesecondnight, towardbothregions(4inNGC6334Iand7inNGC6334I(N)). theCH CN(5 −4 )transitionsat∼91.98GHzweretargeted 3 K K Furthermore, Hunter et al. (in prep.) identified an additional averagingtwopolarizationstoachievebettersignal-to-noisera- SiOoutflowinNGC6334I(N)thathasitsorientationinnorth- tio.Formoredetailsonthespectrallines,seeTable1.Agood east south-west direction, approximately perpendicular to the uv-coveragewas obtained throughregular switching between onepreviouslyreportedbyMegeath&Tieftrunk(1999). both sources and the gain calibrators 1742-289 and 1759-39. Here we present 3.4mm continuum and HCN/CH CN ThechannelseparationoftheHCNobservationswas0.25MHz 3 spectral line observations obtained with the new 3mm facil- (∼0.85kms−1)andslightlyworsewith0.5MHz(∼1.63kms−1) ityattheATCA.Theseobservationsshedlightontheoutflow forCH CNbecausewehadtocoverabroaderbandwidthdue 3 anddensegaspropertiesofbothregionsaswellasonthecon- to the several K-components.Flux calibration was performed tinuum emission from the embedded protostellar objects and with observations of Uranus and is estimated to be accurate theUCHregion. within20%.TheprimarybeamoftheATCAattheobserving frequency is 36′′ (FWHM). The data were reduced with the MIRIAD package. Applying different weightings for the line 2. Observations and continuumdata – mostly uniformweightingforthe com- The two regions NGC6334I & I(N) were observed in May pactcontinuumandCH CNemissionandnaturalforthemore 3 2006 during two nights with the ATCA in the H214 con- extendedHCNemission–thesynthesizedbeamsvarybetween figuration that results in projected baselines between 14 and thedifferentmaps.Theachievedspatialresolutionand1σrms 73kλat88GHz.TheweatherconditionsatNarrabriwereex- valuesaregiveninTable2. Beutheretal.:ATCA3mmobservationsofNGC6334IandI(N) 3 Table2.Synthesizedbeamsθandrms nificant flux at 3.4mm (e.g.,Beutheretal. 2007a), we refrain fromfurtheranalysisofthatfeature. Line θ 1σrms [′′] [mJy] beam 3.1.2. ThedensegasobservedinCH CN(5 −4 ) NGC6334I 3 K K 3.4mmcont. 2.3′′×1.6′′ 23 In contrast to the 3.4mm continuum emission, the CH CN CH CN(5 −4 ),[-12,2]km/sa 2.2′′×1.6′′ 7 3 3 4 4 emission distribution shows the typical double-peaked CH CN(5 −4 ),1.7km/sb 2.2′′×1.6′′ 11 3 4 4 morphology known from the previous NH observations HCN(1–0),[-25,-19]km/sa 2.6′′×1.8′′ 19 3 (Beutheretal.2005,2007b).ThetwoCH CNpeaksappearto HCN(1–0),[5,19]km/sa 2.6′′×1.8′′ 16 3 HCN(1–0),1.7km/sb 2.5′′×1.8′′ 33 be associated with the two strongest mm continuum sources (mm1 and mm2) in the region (Hunteretal. 2006). Figure 3 NGC6334I(N) 3.4mmcont. 2.6′′×1.8′′ 2.6 presentsthefullCH3CN(5K−4K)spectraextractedtowardthe CH CN(5 −4 )a,[-12,8]km/s 2.2′′×1.6′′ 4.9 two peak positions, and clearly all five K-components up to 3 k=0,1 k=0,1 CH CN(5 −4 ),1.7km/sb 2.2′′×1.6′′ 8.4 K = 4 with E /k = 128K are well detected.Table 3 lists the 3 4 4 u HCN(1–0),[-16,-6]km/sa 2.5′′×1.8′′ 19 fittedlineparametersofthespectra. HCN(1–0),[4,12]km/sa 2.5′′×1.8′′ 15 HCN(1–0),2.0km/sb 2.5′′×1.8′′ 26 aIntegratedvelocityregime. bChannelmapswiththegivenvelocityresolution. 3. ResultsandDiscussion All of the targettedspectral lines with excitation levels above ground,E /k, between4 and128K (Table 1) and the 3.4mm u continuum emission were detected and mapped toward both targetregions(Figs.1&2).WhileCH CNisatypicalhotcore 3 moleculeandonlydetectedtowardthecentralwarmprotostars, HCNexhibitssignificantlymoreextendedemissionandshows astrongassociationwiththevariousmolecularoutflowsinthe tworegions.Thismaybeconsideredsurprisingsincethecrit- ical density of HCN(1–0) is even larger than that of CH CN. 3 The 3.4mm continuumemission is of differentorigin in both regions:while it is largely due to the free-freeemission from the UCH region in NGC6334I, toward the northern region NGC6334I(N)thecontinuumemissionstemsmainlyfromthe dustinthevicinityoftheembeddedprotostars.Inthefollowing wewilloutlinethecharacteristicsofbothregionsseparately. 3.1.NGC6334I 3.1.1. 3.4mmcontinuum emission Figure 1 presents an overlay outlining the main features of the spectral line and continuum emission toward NGC6334I. Fig.3. CH CN(5 −4 ) spectra(K = 0...4)extractedtoward 3 K K Almostall of the 3.4mmcontinuumemission arises fromthe thetwoCH CNpeakpositionsinNGC6334IshowninFigure UCH region with a comparable morphologyto the previous 3 1. The dotted line in the top panel shows a model spectrum cmcontinuumimagesofthatregion(e.g.,dePreeetal. 1995; createdwithXCLASSatatemperatureof200K. Beutheretal. 2005). In contrast to that, it is only barely de- tectableata3.8σlevelof87mJybeam−1 towardthestrongest 1.4mm continuum, NH and CH CN peak position, mm1 Since the early work by Loren&Mundy (1984) CH CN 3 3 3 in Fig. 1 (Hunteretal. 2006; Beutheretal. 2007b), mark- has often been used as a thermometer to estimate rotation ing the location of the dominating protostar in the region. temperatures of the dense gas via Boltzmann plots assuming Comparingthis3.8σdetectionwiththe1.4mmdata-pointfrom optically thin emission in Local ThermodynamicEquilibrium Hunteretal. (2006) of 2.09Jybeam−1 (the 1.4mm data were (LTE). We tried this approach here as well, however, it re-imagedwiththesamebeam-sizeasthe3.4mmdata),weget failed because CH CN(5 −4 ) is optically thick. Similarly, 3 K K aspectralindexof∼3.4.However,sinceourdetectionisbarely we tried to model the CH CN spectra in LTE using the 3 abovethe 3σ level, and mm1 may well harbora so far unde- XCLASSsupersettotheCLASSsoftwaredevelopedbyPeter tected hypercompactH regionthatcouldcontributestill sig- Schilke (priv. comm., see also Comitoetal. 2005). This soft- 4 Beutheretal.:ATCA3mmobservationsofNGC6334IandI(N) Fig.1. 3.4mm spectral line and continuumemission toward NGC6334I:The grey-scale shows the 3.4mm continuumand the fullcontourspresentthe integratedCH CN(5 −4 ) emission from-12to 2kms−1. Thedashedand dottedcontoursshowthe 3 4 4 blue-andred-shiftedHCN(1–0)emissionwithintegrationranges[-25,-19]and[5,19]kms−1,respectively.Intheleftpanel,the 3.4mmandCH CNemissioniscontouredfrom3σandcontinuein3σsteps(1σ(3.4mm)∼23mJybeam−1and1σ(CH CN)∼ 3 3 7mJybeam−1).Thered-andblue-shiftedHCNcontoursstartat2σandcontinuein1σsteps(1σ(HCN−red)∼19mJybeam−1 and 1σ(HCN−blue) ∼ 16mJybeam−1). The mm sourcesfrom Hunteretal. (2006) are markedby stars. For clarity, the right panelshows the same data but with less contours,and additionalpoitionsfrom variousother observationsare included (iden- tified at the bottom-left).The correspondingreferencesare: SMA mm continuumdata from Hunteretal. (2006), H O masers 2 fromForster&Caswell(1989),CH OHclassmasersfromWalshetal.(1998),OHmasersfromBrooks&Whiteoak(2001), 3 MIR sources from DeBuizeretal. (2002)). The NH (6,6)/(8,6)/(11,9) are not shown but are spatially associated with mm2 3 (Beutheretal.2007b;Walshetal.2007).Thesynthesizedbeamisshownatthebottomrightofeachpanel. ware package uses the line catalogs from JPL and CDMS components,oneidentifiesatrendofincreasingline-width,∆v, (Poynter&Pickett1985;Mu¨lleretal.2001).However,thisap- withincreasingKquantumnumberforK ≥2(Table3).Figure proach failed as well. Fig. 3 shows a model spectrum at T = 4showsthe∆vofCH CN(5 −4 )forK ≥ 2towardthetwo 3 K K 200K overlaidon the CH CN(5 −4 ) towardthe mainmm CH CN peaks in NGC6334I & I(N) plotted versus the level 3 K K 3 peak mm1. While the K = 2,3 componentsstill fit relatively energyaboveground,E /k.Forallfourpositionsthetrendof u well, the model does neither reproduce the K = 0,1 nor the increasing∆v versus E /k is discernable.Thistrend indicates u K = 4 component. While this is partly again an opacity ef- moreinternalmotionsfromthewarmergascomponents(traced fect, it also shows that LTE is not appropriate for a source bythehigherE /klines)whichperhapsoriginatefromthein- u like NGC6334I. As outlined below, different K-levels exhibit ner warm regions close to the protostars. Similar to that, the different line-widths and hence do not trace the same gas bottom-panel of Figure 5 shows the 2nd moment map of the components.Temperaturegradientswithinthesourcesareim- CH CN(5 −4 )line,i.e.,itsline-widthdistribution.Againwe 3 4 4 printed in the spectra further complicatingsingle-temperature seetheline-widthincreasetowardthecenterclosetothemain fits. More sophisticated modeling of the CH CN emission is mmcontinuumpeaks.Differentprocessesmaycausesuchline 3 warranted. Although a few radiative transfer calculations of broadening,e.g.,accretiondiskrotation,infalloroutflowmo- CH CN data exist (e.g., Olmietal. 1993), the sparsely avail- tions.Inparticular,accretiondisksareinterestingcandidatesto 3 ablecollisionaltransitionratesmakesuchcalculationsa diffi- explainsuchobservationalfeatures. culttask(theonlypartlyavailabledataareasmallcompilation To investigate rotation from a potentially embedded mas- by Pei&Zeng 1995b,a). Furthermore, in such hot and dense sive accretion disk, the top-panelof Figure 5 presents the 1st regionsradiativeexcitationstartstomatterwhichishardtoac- momentmapoftheCH CN(5 −4 )line,i.e.,itspeakvelocity countforinanymodelingapproach. 3 4 4 distribution.Althoughthisstructureisonlybarelyresolvedby WhiletheFullWidthHalfMaximum(FWHM)line-widths thesynthesizedbeamof2.5′′×1.8′′wetentativelyidentifyave- ∆v for the two lowest energy lines of the K = 0,1 compo- locitygradientacrossmm1withanapproximatepositionangle nentsarelesscertainbecauseoftheline-blendingbetweenboth (PA)of113±23degreesfromnorth.Thesamevelocitystruc- Beutheretal.:ATCA3mmobservationsofNGC6334IandI(N) 5 Fig.2.3.4 mmspectrallineandcontinuumemissiontowardNGC6334I(N)Thefourpanelsshowthecontinuum(top-left),the integratedCH CN(5 −4 )(K =0,1,top-right),andthered-andblue-shiftedHCN(1–0)emission(bottom-leftandbottom-right 3 K K respectively).Thevelocityregimesaregiveninthefigure.The3.4mmandCH CNemissioniscontouredfrom2σandcontinue 3 in2σsteps(1σ(3.4mm) ∼ 2.6mJybeam−1and1σ(CH CN) ∼ 4.9mJybeam−1).Thered-andblue-shiftedHCNwingcontours 3 start at 2σ and continue in 1σ steps (1σ(HCN−red) ∼ 19mJybeam−1 and 1σ(HCN−blue) ∼ 15mJybeam−1). Negative features caused by insufficient uv-coverage are shown in dashed contours with the same levels as the emission. Markers of variousotherobservationsareidentifiedinthetop-rightpanel(SMAmmdatafromHunteretal.(2006),thesourcesarelabeled inthetop-leftpanel,CH OHclassmaserfromWalshetal.(1998),cmemissionfromCarraletal.(2002)).Thetwolinesoutline 3 theaxesofthetwomolecularoutflowsidentifiedbyMegeath&Tieftrunk(1999)Hunteretal.(inprep.).Thesynthesizedbeams areshownatthebottomrightofeachpanel. ture is descernable in the lower K = 3,2 CH CN transitions tive of a velocity gradient with a PA of ∼ 134+20 degrees 3 −37 (we refrained from analyzing the K = 0,1 lines because of from north.In comparisonto these position angles, the PA of their line-blending).To investigatethis potentialvelocity gra- the previously identified molecular outflow is ∼ 46 degrees dientinmoredetail,wefittedthepeakpositionsofeachinde- from north (Bachiller&Cernicharo 1990), which is approxi- pendentspectralchannel(Fig.6)whichshouldallowustoin- matelyperpendiculartothatfoundfromourmeasurements,es- creasetheresolvingpowertoapproximately0.5HPBW/(S/N), pecially those from the highest-spatial-resolutionposition fit- whereHPBWequalsthesynthesizedbeamandS/Nthesignal- ting(Fig.6). to-noise ratio (Reidetal. 1988). Similar to the moment map, Based on these findings, we went back to the previous the case is not clear-cut,butnevertheless,the data are indica- NH (1,1) to (6,6) observations (Beutheretal. 2005, 2007b) 3 6 Beutheretal.:ATCA3mmobservationsofNGC6334IandI(N) Table3.FittedCH CN(5 −4 )lineparameters 3 K K Line S ∆vb peak [K] [km/s] NGC6334I,mm1 CH CN(5 −4 )a 24.1 7.0±0.4 3 0 0 CH CN(5 −4 )a 27.8 8.3±0.4 3 1 1 CH CN(5 −4 ) 29.9 7.5±0.2 3 2 2 CH CN(5 −4 ) 31.9 7.8±0.03 3 3 3 CH CN(5 −4 ) 18.7 8.5±0.1 3 4 4 NGC6334I,mm2 CH CN(5 −4 )a 25.6 5.1±0.3 3 0 0 CH CN(5 −4 )a 26.9 5.3±0.3 3 1 1 CH CN(5 −4 ) 27.3 5.3±0.1 3 2 2 CH CN(5 −4 ) 28.8 5.8±0.1 3 3 3 CH CN(5 −4 ) 11.9 6.9±0.2 3 4 4 NGC6334I(N),mm1 CH CN(5 −4 )a 7.4 8.3±0.6 3 0 0 CH CN(5 −4 )a 6.9 8.4±0.8 3 1 1 CH CN(5 −4 ) 8.7 7.6±0.5 3 2 2 CH CN(5 −4 ) 8.9 7.7±0.2 3 3 3 CH3CN(54−44) 4.4 10.0±0.4 Fig.4. Line-widths ∆v(CH3CN) from the various K- NGC6334I(N),mm2 componentsplottedagainsttheupperenergylevelofeachline. CH CN(5 −4 )a 5.6 6.6±1.0 WeomittedtheK =0,1linesbecauseofthelineblending.The 3 0 0 CH3CN(51−41)a 5.7 3.9±0.5 different symbols correspond to 4 different peak positions in CH3CN(52−42) 4.6 6.5±0.5 NGC6334IandI(N)aslabeledintheplot.Thecorresponding CH CN(5 −4 ) 4.5 6.0±0.3 3 3 3 errorsaregiveninTable3. CH CN(5 −4 ) 1.0 10.9±1.6 3 4 4 aThefitstotheK =0,1linesarelessaccuratebecauseofthestrong Adopting the proposed disk scenario, we can estimate lineblendingofbothcomponents. bFullWidthHalfMaximum(FWHM) the approximate rotationally supported binding mass Mrot as- suming equilibrium between the centrifugaland gravitational forcesattheouterradiusofthedisk.Thenweget δv2r searchingforsimilar signaturesin these data. While the main Mrot = (1) G hyperfine lines as well as the satellite lines of NH (1,1) are 3 ⇒ M [M ] = 1.1310−3×δv2[km/s]×r[AU] (2) dominated by the large-scale velocity gradient over the two rot ⊙ main cores (e.g., Fig. 13 in Beutheretal. 2005), the satellite r is the disk radius, and δv the Half Width Zero Intensity linesofthe(J,K)lineswithJ,K≥3areoverlappinganddifficult (HWZI)ofthespectralline,approximately5.1kms−1(halfthe toimage.However,thesatellitehyperfinelinesoftheNH3(2,2) velocity range shown in Fig. 6). Equations 1 & 2 have to be transitionexhibit,in additiontothe velocitygradientoverthe dividedbysin2(i)whereiistheunknowninclinationanglebe- two cores, a second velocity gradient across mm1, again ap- tween the disk plane and the plane of the sky (i = 90◦ for an proximately perpendicular to the large-scale outflow (Fig. 7). edge-onsystem).WiththegivenvalueswecanestimateM to rot Theconfirmationofthisvelocitygradient,firstidentifiedinthe ∼ 8/(sin2(i))M .Thisisofthesameorderasthemassderived ⊙ CH3CN(54−44)line,nowintheloweropacitysatellitehyper- fromthemmcontinuumemission(Hunteretal.2006)whichis fineNH3(2,2)linesupportsitsgeneralcredibility. assumedtostemlargelyfromthedisk/envelopesystem. Sincewedonotresolvewellthesubstructureofthatveloc- How do these masses correspond to the mass and lumi- ity gradient to better investigate the kinematics (e.g., is there nosity of the central embedded source? While the bolometric anyKeplerianmotionpresent?)thedatadonotallowtheclaim luminosity of NGC6334I is estimated to be ≤ 2.6 × 105L ⊙ that a massive accretion disk was detected. For example, an (Sandell 2000), approximately0.32×105L are attributed to ⊙ unresolveddouble-sourcewithdifferentline-of-sightvelocities the UCH region (based on the Lyman continuum flux pre- could produce a similar signature (e.g., Broganetal. 2007). sentedindePreeetal.1995).Oneshouldkeepinmindthatthis Nevertheless, these observations are suggestive of a rotating valuemaybealowerlimitsincedustcouldabsorbasignificant structureperpendiculartothemolecularoutflowwithinapro- fraction of the uv-photons(Kurtzetal. 1994). The remaining jected diameter of ∼ 0.33′′ derived from Fig. 6. At the given ≤ 2.3×105L hastobeduetothevarioussourcesassociated ⊙ distanceof1.7kpc,thiscorrespondstoaradiusoftherotating withthehotmolecularcore.Sincetheassociatedmid-infrared structure of ∼280AU. This scale fits well to the sizes of the sourcetothewest(Fig.1)isofrelativelylowluminosity(only accretiondiskssimulatedrecentlyvia3-dimensionalradiative- 67L , DeBuizeretal. 2002), it is likely that most of the lu- ⊙ transferhydrodynamiccalculationsbyKrumholzetal.(2007). minositystemsfromthetwomaincontinuumandspectralline Beutheretal.:ATCA3mmobservationsofNGC6334IandI(N) 7 Fig.6. Positionaloffsets aroundmm1ofthe differentCH CN 3 (5 −4 )velocitychannelsderivedviaGaussianfitsofthepeak 4 4 emissionineachseparatechannel.Theerror-barsarethenom- inal1σ errorsfromthe fits, andthe numberslabelthe central Fig.5. Momentmaps of CH CN(5 −4 ) toward NGC6334I. velocityforeachposition. 3 4 4 Thetoppanelshowsthe1stmoment(peakvelocities)andthe bottom panel the 2nd moment (line widths). The stars mark the positions of the two main mm continuum sources from 3.1.3. ThemolecularoutflowobservedinHCN(1–0) Hunteretal. (2006). In addition to the velocity gradientfrom The blue- and red-shifted HCN(1–0) emission in Figure between mm1 and mm2, the 1st moment map exhibits a 2nd 1 shows high-velocity emission associated with the velocity gradientaround mm1 in north-westsouth-east direc- molecular outflow in north-east south-western direction tion(approximatelyperpendiculartothelarge-scaleoutflow). (Bachiller&Cernicharo1990;Leurinietal.2006).ThePA of the emission is not exactly the 45 degrees derived previously from the single-dish CO observations but it is closer to 65 degrees.However,suchadiscrepancyisnotnecessarilyasur- prise if one considers the missing flux problem we encounter peaks.In the extremecase, splitting the luminositysimply by intheHCNobservations.Figure8(bottompanel)presentsthe two, it still implies that about ≤ 1.15×105L emanate from ⊙ HCN spectrum extracted toward the mm1 peak position, and the mm1 region. Assuming a ZAMS star, this correspondsto while we see well the blue- and red-shifted emission, nearly an embedded star of ≤30M⊙. In the above adopted accretion allthefluxaroundthev of∼−7.6kms−1isfilteredout.This lsr diskscenario,thiswouldimplyaninclinationangleibetween is also the reasonwhywe donotsee HCN emission fromthe the disk plane and the plane of the sky of approximately 30 hot core itself, which is prominentin CH CN. Therefore, we 3 degrees.However,withthelargeuncertaintiesintherotational just see some selectively chosen part of the outflow in HCN massandthecentralobjectmassestimate,suchaninclination thatcouldforexamplebeassociatedwiththelimb-brightened angleestimateshouldnotbetakenatfacevalue,butonlygives cavity walls of the outflow which would explain the different a roughidea thatthesystem isneitheredge-norface-on.The apparentPA of the image. Figure 9 shows a position-velocity comparable mass derived from the mm continuum emission diagramcenteredat the main mm emission andCH CN peak 3 (Hunteretal. 2006) indicates that a significant fractionof the mm1alongtheapparentaxisoftheHCNemission.Againwe totalsystemmassstemsfromtheproposedaccretiondiskand find no emission around the systemic v but going to higher lsr envelope.Thisimpliesthattherotatingstructureisunlikelyin velocities, the blue- and red-shifted gas exhibits increasing Keplerianmotionbutthatitmaybeaself-gravitatingstructure velocity with increasing distance from the center, closely andpotentialsiteofongoingsub-fragmentation(seealsocom- resembling the typical Hubble-law of molecular outflows parable analytic calculations and hydro-simulations by, e.g., (e.g., Leeetal. 2001). Such Hubble-law like behavior can Kratter&Matzner2006andKrumholzetal.(2007)). be explained on the smallest jet-scales close to the protostar However, on a cautionary note, one has to keep in mind by the decreasing gravitational potential of the central star, thatthedataarenotconclusiveastowhetherwereallyseedisk however, on the larger scales we observe here this effect signaturesorwhethertheobservedgradientmaybecausedby gets negligible and the Hubble-law of molecular outflows othermotions,e.g.,anunresolveddouble-source. is explained by a density gradient decreasing with distance 8 Beutheretal.:ATCA3mmobservationsofNGC6334IandI(N) Fig.8. HCN(1–0) spectra in NGC6334I. The top spectrum shows the strong absorption toward the 3.4mm continuum peakthatcoincideswiththeUCHregionpreviouslyobserved at cm wavelengths. The bottom spectrum is extracted toward the main CH CN peak thatcoincideswith the main mm con- 3 tinuumpeakSMA1byHunteretal.(2006). CO molecular outflow (Beutheretal. 2007b). These features Fig.7. First momentmap (top) and position velocity diagram were interpreted as indicative of this outflow being driven by (bottom)ofthethemostblue-shiftedsatellitehyperfinelineof a sourceassociatedwith mm2.Whilethe caseisnotclear-cut NH (2,2) toward NGC6334I. The data are re-examined from withtwoindependentanddifferentoutflowdriverindications, 3 Beutheretal.(2005).Thepositionvelocitydigramiscentered itisalsopossiblethatwearewitnessingtwodifferentoutflows on mm1 with a position angle of 136 degrees from north. In observed in HCN and CO that may emanate from mm1 and additiontothevelocitygradientsbetweenmm1andmm2,the mm2, respectively. In this scenario, mm1 could be the driver dataexhibita2ndvelocitygradientaroundmm1innorth-west of a potentially denser and younger outflow that is better de- south-eastdirection(approximatelyperpendiculartothelarge- tected in HCN, whereasmm2maybe the driverof thelarger, scaleoutflow).Thesynthesizedbeamis2.8′′×2.2′′. possiblyolderoutflowobservedinCO.Futureobservationsat higherangularresolutionand/orwithdifferentoutflowtracers arerequiredtoassessthevalidityofthisscenario. combined with the continuous(or episodic) driving of the jet that constantly induces energy in the outflow (e.g., Shuetal. 3.1.4. AbsorptiontowardtheUCH region 1991;Smithetal.1997;Downes&Ray1999). Figure 10 (top panels) presents the 1st and 2nd moment Another interesting feature of the HCN data is that we see map(intensity-weightedpeakvelocityandline-widthdistribu- strong absorption in the direction of the UCH region, simi- tions) of the HCN(1–0)emission. It is interesting to note that larto thepreviouslyobservedabsorptioninCH OHandNH 3 3 the line-width distribution clearly peaks toward the strongest (Beutheretal. 2005).Figure8(toppanel)showstheHCN(1– mm continuum and molecular line source mm1 and that the 0) spectrumextractedtowardthe 3.4mmcontinuumpeakpo- large-scale HCN velocity gradient is approximately centered sition. Fitting the HCN(1–0)line-width, we take into account toward that position as well. This is indicative of a scenario itshyperfinestructureconsistingofthreelines(F = 0−1,2− which puts the driving source of the molecular outflow ob- 1,1−1) with relative intensities of 1:5:3 in the optically thin served in HCN at this position. However, previous NH (6,6) limitandvelocityshiftsof-7.1,0and4.9kms−1,respectively 3 maserobservationsshowedthemaserpeakpositionbeingasso- (Poynter&Pickett 1985). The line-width ∆v of that absorp- ciatedwithmm2withadditionalfeaturesalongtheaxisofthe tion feature is 15.4kms−1, much broader than the previously Beutheretal.:ATCA3mmobservationsofNGC6334IandI(N) 9 Fig.10.MomentmapsofHCN(1–0)towardNGC6334I.Theleftandrightcolumnspresentthefirstandsecondmoments(peak velocitiesandlinewidths),respectively.Thetop-rowshowsthemomentsfortheemissionoftheoutflowandambientgas,whereas thebottom-rowpresentsthecorrespondingparametersfortheabsorptionfeaturestowardtheUCHregion.Thecontourspresent the3.4mmcontinuumemissionin3σcontourlevels(3σ∼69mJybeam−1).ThemarkersarethesameasinFig.1. observed line-widths between 1.7 and 2.1kms−1 for CH OH Table4.Lineparametersforallabsorptionlines 3 and between 1.3 and 1.9kms−1 for NH (1,1) and NH (2,2) 3 3 (Beutheretal.2005).Table4liststheobservedline-widthsas Line E /k n ∆v u crit well as the upper level excitation temperatures E /k and the [K] [cm−3] [km/s] u critical densities ncrit for all observed absorption lines. The NH3(1,1) 24 2.1e3 1.4±0.4 other extreme of the line-widthsdistribution is the line-width NH3(2,2) 65 2.2e3 2.0±0.3 oftheionizedgasobservedintheH76αline∆v = 32.0kms−1 CH3OH(32,1−31,2) 36 1.0e4 1.8±0.1 CH OH(4 −4 ) 45 1.2e4 2.1±0.1 (dePreeetal.1995). 3 2,2 1,3 CH OH(2 −2 ) 29 0.8e4 1.7±0.1 3 2,0 1,1 HCN(1–0) 4 2.6e6 15.4±0.3 TheNH andCH OHdataarefromBeutheretal.(2005).Thelisted 3 3 parameters are the upper level excitation energies E /k, the critical u InspectingthevaluesinTable4onecandiscerntwotrends: densities n calculated at 60K and theobserved line-widths of the crit thefirstisthedistinctionbetweenthebroadline-widthandhigh absorptionlinestowardtheUCHregionpeakposition. criticaldensity line HCN versusthe smallline-widthand low criticaldensitylinesfromNH andCH OH.Inaddition,within 3 3 Therearedifferentpossibilitiestoexplainsuchtrends:The thelowcriticaldensitymolecules,onecantentativelyidentifya pictureofanexpandingUCHregion1(Beutheretal.2005)in correlationbetweenincreasingE /kandincreasingline-width. u Althoughjudgingfromtheerrorsthiscorrelationislessclear, 1 IncontrasttothepreviousCH OHandNH absorptionlinesthat 3 3 itremainssuggestive. showedadditional blue-shiftedemissionindicativeof expanding gas 10 Beutheretal.:ATCA3mmobservationsofNGC6334IandI(N) differencesarecausedbytheexpandingUCHregion,possibly inafashioncomparabletothatproposedabove.Nevertheless, without knowing the optical depth of the HCN(1–0) line we cannotdistinguishbetweenthetwoscenarios. 3.2.NGC6334I(N) 3.2.1. Millimetercontinuum emission Figure2(top-leftpanel)presentsthe3.4mmcontinuumemis- sion toward NGC6334I(N). In contrast to NGC6334I, where we only see the UCH region in the 3.4mm continuum, to- ward NGC6334I(N) we detect four out of seven previously identified protostellar 1.4mm dust continuum condensations (Hunteretal. 2006) above a 5σ level of 13mJybeam−1. The peak fluxes S (3.4mm) of the four sources, mm1 to mm3 peak and mm6, are listed in Table 5. Emission features below 5σ are not considered further. The 3.4mm peak associated with Fig.9. Position-velocity diagram of HCN(1–0) in NGC6334I mm1 is the strongest and shows an additional extension to- along the main outflow axis. The cut is centered on the main wardthenorth-western1.4mmsourcemm5.Incontrasttothat, CH3CN peak at positional offset −0.3′′/4.5′′ with a position the1.4mmpeakmm4,whichisassociatedwithcmcontinuum angleof65◦fromnorth-to-east.Emissionatthevelocityofrest andCH OHclassmaseremission,isnotdetectedinourdata 3 around−7.6kms−1islargelyfilteredout. abovethe5σlevel. Forabettercomparisonofthe3.4mmfluxeswiththeprevi- ouslyobserved1.4mmobservations,were-imagedthe1.4mm itssurroundingenvelopeimpliesthatclosertotheUCHregion datawithexactlythesamesynthesizedbeamof2.6′′×1.8′′(po- surface the molecular densities and temperatures are higher sition angle of 84 degreesfrom north). Figure 11 presents an than further outside. Therefore, spectral lines tracing higher overlayoftheSMA1.4mmdatawiththeATCA3.4mmdata densities around an expanding UCH region are expected to atthissamespatialresolution.Whilemm1andmm6areclearly exhibitbroaderline-widthsbecausetheiremittinggasismore separatedinbothwavelengthbands,itisinterestingtonotethat directlyimpactedthanthelower-densitymediumfurtherout.A mm2andmm3mergeinthe1.4mmimageatthereducedlower similarexplanationcouldholdforthedifferentexcitationtem- spatial resolution. The corresponding 1.4mm peak fluxes are peratureregimesaswell.Apointofcautionisthattheopacity listedinTable5.Forthewellseparatedsourcesmm1andmm6 ofHCNisprobablyonetotwoordersofmagnitudelargerthan wecanestimatethespectralindicesαbetweenbothbandsnow, thatof,e.g.,NH .WhilewedonotdetectanyNH satellitehy- 3 3 thederivedvaluesare3.7and3.1,respectively(Table5). perfinestructurelinesin absorptionimplyinglow NH opaci- 3 RecentlyRodr´ıguezetal.(2007)showedthatthecmemis- ties,wecannotinferthatexactlyforHCN.IftheHCN(1–0)op- sion from NGC6334I(N) is caused by free-free emission, ticaldepthwerethathighthatitcouldnottracethedensegas regions close to the expanding UCH region, then the above whereas shortward of 7mm wavelength the spectral energy distribution is dominated by dust continuum emission. In the picture could hardly hold. Another explanation is based on Rayleigh-Jeanslimit,thefluxS scaleswithS ∝ ν2+β whereβ the enhancementof HCN in the molecular outflow. Although theoutflowdoesnotemanatefromtheUCHregionbutfrom isthedustopacityindex.Withthemeasuredspectralindexα, wehavedustopacityindicesof1.7and1.1formm1andmm6, theneighboringmmcontinuumsource(s),dePreeetal.(1995) respectively.While the β value of mm1 is consistent with of- found a velocity gradientin the ionized gas similar to that of ten observed values between 1.5 and 2, it is lower for mm6. the molecular outflow. They suggest that the molecular out- Although the synthesized beams of both maps are the same, flow(s)intheregionmaywelldisturbthevelocityfieldofthe UCHregionandproducethevelocitygradientthisway.Since theuv-coveragewasnotduringtheobservations.Furthermore, thecontinuummapsshowa morepeakedmorphologytoward HCNisknowntobeabundantinmolecularoutflows(seealso mm1 than toward mm6. Thus, it is possible that mm6 has IRAS18566+0408,Zhangetal.2007,andIRAS20126+4104, moreextendedstructureandthatthismaybesampledbetterby Liu et al. in prep.) a similar effect could explain the broad HCN(1–0)line-widthstowardtheUCHregion.However,the the ATCA observations,possibly accountingfor the observed lowervaluesofαandβtowardmm6.However,itisalsofeasi- 1stand2ndmomentmaps(peakvelocitiesandline-widths)of blethatphysicalreasonsareresponsibleforthiseffectbecause theHCN(1–0)absorptionpresentedinFig.10(bottompanels) adecreasingβcanbeattributedtograingrowthincircumstellar show no velocity gradient but a peak of the line-width distri- butionclosetothecenteroftheUCHregion.Thisiscounter- disksaswell(e.g.,Beckwithetal.1990).Nevertheless,inthis scenario, it appears surprising that mm1, which likely drives intuitivefortheoutflowpictureandsuggeststhattheline-width anoutflowandhenceprobablycontainsanaccretiondisk,has aroundtheUCHregion(Beutheretal.2005),theHCNdatadonot a value of β close to the interstellar medium values, whereas exhibitanyclearexpansionorinfallsignature. the source mm6, that does not exhibit clear signs of ongo-