Cloud-Cloud Collision Induced Star Formation in IRAS 18223-1243
PPreprint typeset using L A TEX style emulateapj v. 5/2/11
CLOUD-CLOUD COLLISION INDUCED STAR FORMATION IN IRAS 18223-1243
L. K. Dewangan , D. K. Ojha , I. Zinchenko , and T. Baug ABSTRACTIn the direction of l = 17 ◦ .6 – 19 ◦ , the star-forming sites Sh 2-53 and IRAS 18223-1243 are promi-nently observed, and seem to be physically detached from each other. Sh 2-53 has been investigated atthe junction of the molecular filaments, while a larger-scale environment of IRAS 18223-1243 remainsunexplored. The goal of this paper is to investigate the star formation processes in the IRAS site (area ∼ ◦ .4 × ◦ .4). Based on the GRS CO line data, two molecular clouds, peaking at velocities of 45and 51 km s − , are found. In the position-velocity plots, a relatively weak CO emission is detectedat intermediate velocities (i.e. 47.5–49.5 km s − ) between these two clouds, illustrating a link betweentwo parallel elongated velocity structures. These clouds are physically connected in both space andvelocity. The MAGPIS data at 20 cm trace free-free continuum emission toward the IRAS 18223-1243source. Using the Spitzer and UKIDSS photometric data, we have identified infrared-excess youngstellar objects (YSOs), and have observed their groups toward the intersection zones of the clouds.IRAS 18223-1243 is also spatially seen at an interface of the clouds. Considering these observationalfindings, we propose the onset of the collision of two clouds in the IRAS site about 1 Myr ago, whichtriggered the birth of massive star(s) and the YSO groups. A non-uniform distribution of the GPIPSH-band starlight mean polarization angles is also observed toward the colliding interfaces, indicatingthe impact of the collision on the magnetic field morphology.
Subject headings: dust, extinction – HII regions – ISM: clouds – ISM: individual object (IRAS 18223-1243) – stars: formation – stars: pre-main sequence INTRODUCTION
Massive OB-stars ( ≥ (cid:12) ) are the major and ultimatesources of mechanical and radiative energy in galaxies,which are produced from their birth until their death.The understanding of the formation processes of suchstars is still one of the outstanding topics in star forma-tion research (e.g. Zinnecker & Yorke 2007; Tan et al.2014). In recent years, to explain the origin of massivestars and young clusters, a cloud-cloud collision (CCC)process has been proposed as an interesting alternativeagainst the existing competing theories of massive starformation (i.e. “Turbulent core accretion” and “Compet-itive accretion”). One can find more details about theseprocesses in Zinnecker & Yorke (2007) and Tan et al.(2014). Various numerical studies concerning the CCCsuggested that the colliding molecular gas can producedense and massive cloud cores in the shock-compressedinterface, which could provide the conditions needed formassive star formation (e.g. Habe & Ohta 1992; Anath-pindika 2010; Inoue & Fukui 2013; Takahira et al. 2014,2018; Haworth et al. 2015a,b; Torii et al. 2017a; Bisbaset al. 2017). The intense star formation activity nearthe collision interface is predicted in the CCC process.To date, more than 20 star-forming regions are known,where the observational evidences for the CCC process [email protected] Physical Research Laboratory, Navrangpura, Ahmedabad -380 009, India. Department of Astronomy and Astrophysics, Tata Instituteof Fundamental Research, Homi Bhabha Road, Mumbai 400 005,India. Institute of Applied Physics of the Russian Academy of Sci-ences, 46 Ulyanov st., Nizhny Novgorod 603950, Russia. Kavli Institute for Astronomy and Astrophysics, Peking Uni-versity, 5 Yiheyuan Road, Haidian District, Beijing 100871, P.R. China. have been reported (e.g. Furukawa et al. 2009; Ohama etal. 2010, 2017a,b; Fukui et al. 2014, 2016, 2018; Baug etal. 2016; Dewangan 2017; Dewangan & Ojha 2017; De-wangan et al. 2017; Fujita et al. 2017; Torii et al. 2017a,b;Hayashi et al. 2018; Kohno et al. 2018; Sano et al. 2018).Infrared-dark clouds (IRDCs) are often investigated aselongated filamentary clouds, and also harbor the earlyphases of massive OB-stars (see reviews by Bergin &Tafalla 2007; Andr´e et al. 2014, and references therein).It implies that such elongated clouds can be consideredas the potential sites to probe the formation mechanismsof massive stars (e.g. Rathborne et al. 2006, 2007, 2010;Bergin & Tafalla 2007; Zhang et al. 2017).In this paper, we have selected a known filamentarystructure IRDC 18223 (e.g. Tackenberg et al. 2014, andsee Figure 1 in their paper) containing a massive star-forming region IRAS 18223-1243, and have performed adetailed multi-wavelength study of a large-scale environ-ment around this IRAS site. IRAS 18223-1243 is a highmass protostellar object (HMPO) candidate (e.g. Srid-haran et al. 2002), and is located at a distance of 3.5kpc (Sridharan et al. 2002; Beuther et al. 2002, 2015;Beuther & Steinacker 2007; Fallscheer et al. 2009; Lu etal. 2014; Tackenberg et al. 2014). The IRDC 18223 hasbeen studied using the multi-scale and multi-wavelengthdata including the Submillimeter Array (SMA), the VeryLarge Array (VLA), the Nobeyama 45 m telescope, andthe IRAM Plateau de Bure interferometer (PdBI) facil-ities (e.g. Fallscheer et al. 2009; Beuther & Steinacker2007; Beuther et al. 2015). Using the VLA NH obser-vations, Lu et al. (2014) reported that IRAS 18223-1243is filamentary in the north-south direction (see Figure 1in their paper). They also measured a radial velocityof the molecular gas associated with the IRAS site (i.e.44.9 km s − ) similar to that of the ionized region U18.66- a r X i v : . [ a s t r o - ph . GA ] M a y L. K. Dewangan et al.0.06 (i.e. 44.1 km s − ; Anderson & Bania 2009). Us-ing the molecular line and millimeter continuum data,Beuther et al. (2015) identified at least 12 dusty cores(M core ∼ (cid:12) ) in the 4 pc long filamentary cloudIRDC 18223 (see Figure 3 in their paper), and suggestedthat the IRDC 18223 is an excellent example of a mas-sive gas filament. They also suggested that these embed-ded cores could be at different evolutionary stages of themassive star formation. In one of these cores, Fallscheeret al. (2009) also previously reported the presence of amolecular outflow and evidence for a large rotating ob-ject perpendicular to the outflow (see also Figure 1 inTackenberg et al. 2014). Using the N H + spectral linedata, Beuther et al. (2015) identified a gradient in ve-locity perpendicular to the main filament, however novelocity gradient was found along the axis of the fila-ment. Using the lower-density gas tracers (such as, [CI]and C O), these authors found red- and blue-shifted ve-locity structures on scales around 60 (cid:48)(cid:48) east and west ofthe IRDC 18223 filament. This result was interpreted asa signature of the large-scale cloud and the smaller-scalefilament being kinematically coupled. Together, theseprevious studies suggest the ongoing massive star forma-tion activity in the IRDC 18223.One can note that the earlier works were mainly fo-cused toward the 4 pc long filamentary cloud IRDC 18223containing the HMPO candidate. The investigation of alarge-scale environment (more than 10 pc) around theIRAS 18223-1243 site is yet to be carried out despitethe presence of numerous observational data sets. Usinga multi-wavelength observational approach, our presentwork focuses to understand the physical environment andstar formation processes around IRAS 18223-1243. Thepaper also includes a detailed analysis of CO line datato study the kinematics of the structures embedded inthe IRAS site.The paper is arranged as follows. Section 2 briefsthe multi-wavelength data sets. In Section 3, we detailour findings derived through a multi-wavelength obser-vational approach. In Section 4, we discuss the ongoingphysical processes in our selected IRAS site. Finally, weconclude the paper in Section 5. DATA SETS
In this paper, the observational data sets were retrievedfrom various publicly available surveys (see Table 1).Elaborative details of these data sets and their reduc-tion processes can be found in Dewangan (2017). RESULTS
Large scale physical environment of IRAS18223-1243
In this section, we present multi-wavelength images ofIRAS 18223-1243 to investigate the large-scale morphol-ogy.The star-forming sites Sh 2-53 and IRAS 18223-1243are spatially seen in the direction of l = 17 ◦ .6 – 19 ◦ ; b = 0 ◦ .13 – − ◦ .8. Figure 1a presents the Herschel µ m image of an area ∼ ◦ .42 × ◦ .42 containing thesetwo sites, displaying the distribution of cold dust emis-sion. In the site Sh 2-53 (see a dashed box in Figure 1a),ongoing massive star formation activities have been re-ported by Baug et al. (2018) (see also Ohama et al. 2017a, and references therein). Using the GRS CO line data(having a velocity resolution of 0.21 km s − , an angu-lar resolution of 45 (cid:48)(cid:48) , and a typical rms sensitivity (1 σ )of ≈ .
13 K; Jackson et al. 2006), they reported thatthe molecular gas in the site Sh 2-53 is traced in a ve-locity range of 37–60 km s − . Figure 1b presents anintegrated CO intensity map of the area highlightedby a solid box in Figure 1a, and the emission is inte-grated over a velocity range of 37–60 km s − . The IRASsite is also observed in the molecular map. Figure 1cshows a position-velocity (p-v) plot along the axis in thedirection of these two sites. The velocity structures to-ward these two sites appear different, indicating the onsetof different physical processes. Furthermore, the veloc-ity structure at the position ( l = 18 ◦ .242; b = − ◦ .293)shows a discontinuity in the molecular emission, suggest-ing that Sh 2-53 could be physically detached from IRAS18223-1243. Note that the site Sh 2-53 is extensively ex-plored by Baug et al. (2018). These authors found thesite Sh 2-53 at the junction of molecular filaments (i.e.“hub-filament” system), and the flow of gas was also in-vestigated toward the junction. However, a larger-scalephysical environment of IRAS 18223-1243 is yet to bestudied. In this work, our analysis is mainly performedfor a field of ∼ ◦ .4 × ◦ .4 ( ∼
24 pc ×
24 pc; centered at l = 18 ◦ .611; b = − ◦ .102) around the IRAS 18223-1243site (see a dotted-dashed box in Figure 1b).Figures 2a, 2b, and 2c present the Herschel µ m,ATLASGAL 870 µ m, and Spitzer µ m images of IRAS18223-1243, respectively. The MAGPIS 20 cm contin-uum emission is also overlaid on the Herschel µ mand the Spitzer µ m images (see Figures 2a and 2c).Figure 2c shows a zoomed-in view of the 8.0 µ m imagetoward the IRAS 18223-1243 position. The ionized emis-sion/H ii region traced in the MAGPIS map is observedtoward the IRAS 18223-1243 (see Figures 2a and 2c).In the sub-millimeter (sub-mm) 250 and 870 µ m maps,several cold condensations and at least two elongated em-bedded filamentary features, with lengths larger than 20pc, are found in our selected target field (see Figures 2aand 2b). These filamentary features are designated asfl1 and fl2. In Figure 2c, the Spitzer µ m image isalso superimposed with the 8.0 µ m emission contour, re-vealing an extended feature containing the IRAS posi-tion. The positions of IRAS 18223-1243 and the 20 cmpeak emission seem to be spatially matched, and are alsofound within the extended 8.0 µ m feature. Furthermore,the IRDC 18223 (length ∼ Spitzer µ m image (see Figure 2c), andis identified against the mid-infrared background emis-sion. In Figure 2c, we have also marked the positionsof 12 dust continuum cores identified by Beuther et al. This publication makes use of molecular line data from theBoston University-FCRAO Galactic Ring Survey (GRS). The GRSis a joint project of Boston University and Five College Radio As-tronomy Observatory, funded by the National Science Foundationunder grants AST-9800334, AST-0098562, and AST-0100793. TheNational Radio Astronomy Observatory is a facility of the NationalScience Foundation operated under cooperative agreement by As-sociated Universities, Inc. The Infrared Processing and Analysis Center / California In-stitute of Technology, funded by NASA and NSF), archival dataobtained with the
Spitzer
Space Telescope (operated by the JetPropulsion Laboratory, California Institute of Technology under acontract with NASA). loud-Cloud collision in IRAS 18223-1243 3
Table 1
A list of multi-wavelength surveys utilized in the present work.
Survey Wavelength(s) Resolution ( (cid:48)(cid:48) ) ReferenceMulti-Array Galactic Plane Imaging Survey (MAGPIS) 20 cm ∼ ∼
45 Jackson et al. (2006)APEX Telescope Large Area Survey of the Galaxy (ATLASGAL) 870 µ m ∼ Herschel
Infrared Galactic Plane Survey (Hi-GAL) 70, 160, 250, 350, 500 µ m ∼ ∼ ∼ ∼ ∼
37 Molinari et al. (2010)
Spitzer
MIPS Inner Galactic Plane Survey (MIPSGAL) 24 µ m ∼ Spitzer
Galactic Legacy Infrared Mid-Plane Survey Extraordinaire (GLIMPSE) 3.6, 4.5, 5.8, 8.0 µ m ∼ ∼ ∼ ∼ µ m ∼ µ m ∼ µ m ∼ (2015), which are spatially distributed toward the IRDC18223. Previously, it was reported that the IRAS sourceis embedded in one of these cores, which can also be in-ferred from Figure 2c. In the sub-mm maps, the 4 pclong filamentary cloud IRDC 18223 appears to be partof the elongated filament fl1. Additionally, Figure 2calso reveals the ionized emission (or an H ii region) to-ward IRAS 18223-1243. Using the equation given in Mat-sakis et al. (1976), we have computed the number of Ly-man continuum photon (N uv ) to be ∼ × s − (orlogN uv ∼ ii region associated with IRAS18223-1243 (see Dewangan et al. 2016, for more details),which corresponds to a single ionizing star of spectraltype B0.5V-B0V (see Table II in Panagia 1973, and alsoTable 1 in Smith et al. 2002). In the analysis, we useda distance of 3.5 kpc and a typical value of the electrontemperature of 10000 K. The integrated flux density andthe radius (R HII ) of the H ii region are estimated to be158.7 mJy and 0.43 pc, respectively. We have also com-puted the dynamical age of the H ii region to be ∼ × yr to take into account a typical value of the ini-tial particle number density of the ambient neutral gas(n = 10 cm − ), the isothermal sound velocity in theionized gas (c s = 11 km s − ; Bisbas et al. 2009), R HII ,and N uv . One can find more details about the analysisin Dewangan (2017). Distribution of molecular gas
In this section, we present the kinematic analysis ofthe GRS CO in the direction of IRAS 18223-1243.In Figure 3, we show the integrated GRS CO(J=1 −
0) velocity channel maps (at intervals of 1 kms − ). The maps cover a velocity range from 37 to 57 kms − , indicating the presence of different molecular com-ponents along the line-of-sight. In the velocity channelmaps, at least two different elongated molecular cloudcomponents are seen at velocity ranges of 44–45 and 49–50 km s − . Figure 4a shows an integrated CO intensitymap of our selected target field around IRAS 18223-1243.In the map, the molecular emission is integrated over avelocity range from 37 to 57 km s − . The elongated mor-phology of the cloud is also seen in the CO intensitymap. Figure 4b presents a p-v plot of CO, which iscomputed along an axis (see a solid line in Figure 4a). Adotted line (in black) is also highlighted in the p-v plotto indicate the position of IRAS 18223-1243. At leasttwo prominent velocity components (at 45 and 51 kms − ; see broken curves in Figure 4b) and a third one (at55 km s − ) are also seen toward the IRAS location. Wefind a relatively weak CO emission between the twoprominent velocity components, which are separated by ∼ − . In the integrated intensity map, we have also marked the areas of different small fields (i.e. ar1 toar6; see boxes in Figure 4a), where the average spectraare obtained. Figures 4c, 4d, 4e, 4f, 4g, and 4h show theaveraged CO spectra toward the small fields ar1, ar2,ar3, ar4, ar5, and ar6, respectively. In the direction oftwo fields, ar2 and ar3, we find at least three velocitypeaks (at 45, 51, and 55 km s − ) in the profiles, whileonly two velocity peaks are observed toward remainingfour fields ar1, ar4, ar5, and ar6. Note that the locationof IRAS 18223-1243 is seen toward the field ar3. Previ-ously, Beuther et al. (2015) also observed three velocitypeaks (at 45, 51, and 55 km s − ) in the C O(2-1) spec-trum (see Figure 8 in their paper). Based on the p-vplot and the CO spectra, Figure 5 presents the spatialdistribution of molecular cloud components at four dif-ferent velocity ranges (i.e. 37–47, 47.5–49.5, 50–53.5, and54–57 km s − ; see also Figure 4b). Note that moleculargas at a velocity range of 47.5–49.5 km s − connects thetwo velocity components at 45 and 51 km s − (see alsoFigure 4b).In Figure 6, we also present the integrated CO in-tensity map, latitude-velocity plot and longitude-velocityplot. Figure 6a shows an integrated intensity map of CO from 37 to 57 km s − , which is the same as shownin Figure 4a. Figures 6b and 6d present the latitude-velocity and longitude-velocity plots of CO emission,respectively. We have also highlighted two velocity peaks(at 45 and 51 km s − ) in these p-v plots, which are sep-arated by ∼ − . The velocity peak at 55 km s − isnot traced in Figures 6b and 6d, but this velocity compo-nent has been observed in Figures 4b, 4d, and 4e. In thispaper, we have not discussed separately the velocity com-ponent at 55 km s − . In Figures 6b and 6d, a relativelyweak CO emission between two velocity peaks (at 45and 51 km s − ) is also seen (see also Figures 4b and 5b).In Figure 6c, we present the spatial distribution of molec-ular gas associated with two molecular cloud componentsat 37–47 and 49–57 km s − . Using the sub-mm maps, wehave investigated the two elongated filaments fl1 and fl2in our selected target field (see Figure 2). A relative com-parison of the infrared features against the distributionof molecular gas reveals that each elongated filament isembedded in a molecular cloud (see Figures 2a and 6c).The filaments fl1 and fl2 are associated with the cloudstraced in the velocity ranges of 37–47 and 49–57 km s − ,respectively.Together, the analysis of the GRS CO line data sug-gests that two molecular clouds are connected in bothspace and velocity. Interestingly, IRAS 18223-1243 isspatially seen at one of the interfaces of these two clouds,where massive star formation is evident. L. K. Dewangan et al.
Temperature and column density maps of IRAS18223-1243
In this section, we discuss the
Herschel temperatureand column density maps of IRAS 18223-1243, whichare derived using the
Herschel µ m data. Moredetails of the procedures for producing these maps canbe learned from Mallick et al. (2015) (see also Dewanganet al. 2015, 2017; Dewangan 2017; Baug et al. 2018).The temperature and column density maps (resolu-tions ∼ (cid:48)(cid:48) ) are shown in Figures 7a and 7b, respectively.The infrared structure seen in the Herschel µ m is welltraced in a temperature range of about 19–22 K in the Herschel temperature map. The
Herschel temperaturemap traces the previously known 4 pc long filamentarycloud IRDC 18223 in a temperature range of about 16–18 K, while the H ii region is depicted in a temperaturerange of about 25-28 K. The embedded infrared structureand the filamentary features are also traced in the col-umn density map, where several condensations are alsoobserved (see Figure 7b). We employed a “ clumpfind ”IDL program (Williams et al. 1994) in the Herschel col-umn density map to identify clumps and to compute theirtotal column densities. This exercise yields a total of27 clumps in our selected target field, which are high-lighted in Figure 7c. The extension of each clump is alsoshown in Figure 7c. Following the procedures describedin Mallick et al. (2015), the mass of each
Herschel clumpis computed. Table 2 provides the IDs referred to theclump, Galactic coordinates (l, b), deconvolved effectiveradius ( R clump ), and clump mass ( M clump ). The clumpmasses vary between 305 and 3700 M (cid:12) .In Figure 8a, we show a two color-composite image de-rived using the Herschel column density map (red) and
Herschel µ m (green) image. The N (H ) map is ex-posed to an edge detection algorithm (i.e. Difference ofGaussian (DoG); see Gonzalez & Woods 2011; Assiratiet al. 2014). The embedded filaments fl1 and fl2 areclearly visible in the composite map. The MAGPIS 20cm emission contours are also overlaid on the compositemap. Figure 8b presents the spatial distribution of twomolecular clouds similar to those shown in Figure 6c.At least two zones of clouds appear to be spatially over-lapped (see arrows in Figure 8b). To further examinethe elongated molecular clouds, in Figure 8c, we displaythe CO emission contour map at [37, 47] km s − witha level of 14 K km s − , which is also overlaid with the CO emission contour at [49, 57] km s − with a level of12 K km s − . We have also estimated the masses of theelongated molecular clouds to be ∼ (cid:12) (for theelongated cloud at [37, 47] km s − ; see Figure 8c) and ∼ (cid:12) (for the elongated cloud at [49, 57] km s − ;see Figure 8c). In the calculation, we have considered anexcitation temperature of 20 K, the ratio of gas to hy-drogen by mass of about 1.36, and the abundance ratio(N(H )/N( CO)) of 7 × . Elaborative details aboutthe molecular mass calculation can be found in Yan etal. (2016) (see also equations 4 and 5 in their paper).Together, based on Figures 8a, 8b, and 8c the distri-bution of the column density and the molecular gas hasenabled us to further visually infer the two filamentarymolecular clouds (see also Section 3.2). Study of embedded young stellar population
Selection of young stellar objects
In this section, using the
Spitzer and UKIDSS photo-metric data, four methods are employed to select youngstellar objects (YSOs) in the selected target field. Wehave been extensively using these four methods to se-lect YSOs (e.g. Dewangan 2017; Dewangan et al. 2017,2018; Baug et al. 2018), which are the
Spitzer color-magnitude scheme (i.e. [3.6] − [24] vs [3.6]; see Fig-ure 9a), four Spitzer µ m bands (see Figure 9b),three Spitzer µ m bands (see Figure 9c), and near-infrared (NIR) color-magnitude scheme (i.e. H − K/K; seeFigure 9d). Elaborative details of these schemes can beobtained from Dewangan et al. (2018).The photometric data at 3.6-8.0 µ m were collectedfrom the GLIMPSE-I Spring ’07 highly reliable photo-metric catalog. We retrieved the photometric magni-tudes of point-like sources at 24 µ m from Gutermuth& Heyer (2015). We also used the photometric HKdata from the UKIDSS GPS sixth archival data release(UKIDSSDR6plus) catalog and the 2MASS. Figure 9ashows a color-magnitude plot ([3.6] − [24]/[3.6]) of 307sources. The Spitzer color-magnitude scheme gives 77YSOs (32 Class I; 14 Flat-spectrum; 31 Class II) and 230Class III sources. In Figure 9a, red circles, red diamonds,and blue triangles indicate Class I, Flat-spectrum, andClass II YSOs, respectively.Figure 9b shows a color-color plot ([5.8] − [8.0] vs[3.6] − [4.5]) of sources. This scheme yields 53 YSOs (20Class I; 33 Class II), and 2 Class III sources. In Fig-ure 9b, red circles and blue triangles represent Class Iand Class II YSOs, respectively.Figure 9c shows a color-color plot ([4.5] − [5.8] vs[3.6] − [4.5]) of sources. Using this scheme, 19 protostarsare identified in our selected region.Figure 9d shows a NIR color-magnitude plot (H − K/K)of sources. We find a color H − K cut-off value (i.e. ∼ − K excess sources. Using this color condition,this scheme yields 186 YSOs (see Figure 9d).These four schemes are not mutually exclusive. There-fore, we have removed common YSOs selected throughdifferent schemes. Finally, our catalog contains 335YSOs in our selected region around the IRAS 18223-1243site. In Figure 10a, these selected YSOs are overlaid onthe
Herschel column density map. We find a large num-ber of YSOs toward the clouds.
Groups of young stellar objects
In this section, we study the surface density of all theselected 335 YSOs to access their individual groups orclusters. The nearest-neighbour (NN) technique has beenemployed to compute surface density map of YSOs (seeGutermuth et al. 2009; Bressert et al. 2010; Dewanganet al. 2018, for more details). This exercise is completedusing a 5 (cid:48)(cid:48) grid and 6 NN at a distance of 3.5 kpc. Fig-ures 10b and 10c show the surface density contours ofYSOs overlaid on the
Herschel column density map andthe integrated CO intensity map, respectively. TheYSOs density contour levels are 2, 3, and 5 YSOs/pc .The positions of Herschel clumps are also shown in Fig-ures 10b and 10c. The groups of YSOs are traced towardthe spatially common zones of two clouds (see arrows inFigure 10c).loud-Cloud collision in IRAS 18223-1243 5 DISCUSSION
The present work provides a more detailed investiga-tion of ongoing physical processes in the IRAS 18223-1243 site. In the previous sections, for the first time,we have reported the presence of two molecular clouds(at 45 and 51 km s − ) toward the IRAS site, which arealso connected in the velocity space at a velocity rangeof 47.5–49.5 km s − (see Figures 4b and 5b). Both themolecular clouds contain elongated filamentary features(lengths >
20 pc), and at least two zones of clouds seemto be spatially overlapped, and IRAS 18223-1243 is alsoobserved toward one of the common sections. Our ra-dio continuum data analysis suggests the existence of atleast a massive star B0.5V as the powering source forthe observed H ii region in the IRAS 18223-1243 site.Hence, the feedback of the massive star (such as, stel-lar wind, ultraviolet radiation, and pressure-driven H ii region) may be one of the possibilities to explain the ob-served velocity separation between the two clouds. Theexpected mechanical luminosity of the stellar wind (L w = 0.5 ˙ M w V w erg s − ) for B0.5V star can be estimatedusing the typical values of the mass-loss rate ( ˙ M w = 2.5 × − M (cid:12) yr − ; Oskinova et al. 2011) and the wind ve-locity (V w = 1000 km s − ; Oskinova et al. 2011). Withthe help of this analysis, we can further estimate the me-chanical energy (E w ) that can be injected by the massiveB0.5V star in a certain time period. We have obtainedthe values of L w and E w to be ≈ × ergs s − and ≈ × ergs (for the time period of 0.5 – 1Myr), respectively. Using the molecular masses of thetwo clouds (see Section 3.3) and their velocity separa-tion (i.e. ∼ − ), we have computed the kinematicenergy of these clouds to be about 2–3.8 × ergs,which is significantly higher than the mechanical energyfrom the massive star, indicating that the velocity sepa-ration between the two clouds cannot be explained by thestellar feedback (e.g. Furukawa et al. 2009). Although,there is always uncertainty involved in the conversionrate from stellar feedback into the kinetic energy of theclouds. Interestingly, in the CCC process, one may com-fortably expect such a large velocity separation betweentwo molecular clouds.There is growing evidence that the CCC process canexplain the presence of clusters of YSOs and massivestar formation activities at the overlapped section of twomolecular clouds. Our molecular data analysis has re-vealed that the two cloud components (at 45 and 51 kms − ) are physically connected in both space and velocity,hinting for a possibility of interaction between them. Inthe velocity space, the relatively weak emission betweenthe two cloud components may be treated as a broadbridge feature, which could give a clue of the existenceof a compressed layer of gas due to the collision betweenthe clouds seen along the line of sight (e.g., Haworth etal. 2015a,b; Torii et al. 2017a). Additionally, the channelmaps of CO trace a possible complementary molecularpair of the two colliding clouds at [43, 44] km s − and[52, 53] km s − . One can find more details of the ob-servational characteristic features of the CCC in Torii etal. (2017a) (see also Baug et al. 2016; Dewangan 2017;Dewangan & Ojha 2017; Dewangan et al. 2017).With the help of the multi-wavelength data sets, we have also investigated the groups of YSOs, the H ii re-gion, and several massive clumps ( M clump ∼ (cid:12) ) toward the overlapped areas of the molecularclouds. Using the 20 cm continuum data, the age of theH ii region is computed to be ∼ × yr. The av-erage life time of YSOs was reported to be ∼ ∼ ∼ − ) was occurred about 1Myr (i.e. 6.2 pc/6 km s − ) ago. Hence, the birth processof groups of YSOs and massive star(s) in the IRAS siteappears to be explained by the collision of two molecularclouds along the line-of-sight.In order to further assess the impact of colliding cloudsin the IRAS site, we have examined the archival H-band(1.6 µ m) linear polarimetric data from the GPIPS DataRelease 2 (i.e. DR2). Note that the starlight polari-metric data allow us to trace the magnetic field direc-tion in the plane-of-sky (POS) parallel to the directionof polarization (e.g. Davis & Greenstein 1951). Fig-ure 11a shows the GPIPS H-band polarization vectorsof 862 background stars overlaid on the
Herschel imageat 250 µ m. To obtain these background stars, we em-ployed conditions on sources with Usage Flag (UF) =1 and P/σ P ≥
3, where P is the polarization percent-age and σ p is the polarimetric uncertainty. The NIRpolarimetric data are useful to infer the large-scale mag-netic field morphology of the cloud. However this dataset may not be able to provide a detailed informationin the direction of the densest part of the cloud, wherevery high value of extinction is expected. Figure 11bdisplays mean polarization vectors to infer the magneticfield morphology in the IRAS site. To generate the meanpolarization vectors, our target field is divided into 11 ×
10 equal sections and a mean polarization value is com-puted using the Q and U Stokes parameters of H-bandsources traced within each division. Here, we have ap-plied the same procedures as given in Dewangan (2017)(see also Dewangan et al. 2018) to analyze the GPIPSpolarimetric data. In Figures 11a and 11b, the length ofa vector represents the degree of polarization, while theangle of a vector shows the polarization galactic positionangle. In Figure 11b, we have compared the distributionof mean polarization angles toward the common zones ofthe two molecular clouds against their other parts. Thiscomparison reveals a non-uniform distribution of meanpolarization angles toward the interfaces of the clouds.Keeping in mind the higher value of the kinematic en-ergy of the clouds against the mechanical energy from themassive star (as mentioned in this section), it is unlikelythat the observed variation in the mean polarization an-gles can be fully explained by the feedback of massivestar. Therefore, the CCC appears to be the major pro-cess responsible for the observed deviation of magneticfield lines toward the interfaces of the clouds. SUMMARY AND CONCLUSIONS This publication makes use of the Galactic Plane InfraredPolarization Survey (GPIPS). The GPIPS was conducted usingthe
Mimir instrument, jointly developed at Boston University andLowell Observatory and supported by NASA, NSF, and the W.M.Keck Foundation.
L. K. Dewangan et al.The present paper focuses to probe the star formationprocesses on a larger-scale (size ∼ ◦ .4 × ◦ .4) aroundthe IRAS 18223-1243 site. The major outcomes of thepaper are given below: • Using the CO line data, two molecular cloud com-ponents, peaking at velocities of 45 and 51 km s − , areidentified toward the IRAS site. In the velocity spaceof CO, these clouds are interconnected by a relativelyweak intermediate velocity emission. • Using the
Herschel and ATLASGAL sub-mm images,at least two elongated filamentary features (havinglengths >
20 pc) are traced. • Each elongated filamentary feature is embedded in amolecular cloud. • The
Herschel temperature map traces the embeddedfilaments (including the H ii region) in a temperaturerange of about 16–28 K. • A total of 27 clumps are found in our selected site,and their mass ranges are 305–3700 M (cid:12) . • Using the
Spitzer and UKIDSS photometric data, starformation activity is investigated toward the clouds,where massive clumps are found. Groups of YSOs arealso observed toward the spatially overlapped zones ofthe clouds. • IRAS 18223-1243 is located at one of the commonzones of the clouds. The VLA MAGPIS radio contin-uum emission at 20 cm is traced toward the IRAS source.Taken together, our observational results are consistentwith the CCC scenario, supporting the onset of the CCCin the IRAS site about 1 Myr ago. Hence, the formationof YSO groups and massive star(s) is influenced by theinteraction of the molecular clouds in the IRAS site.We thank the anonymous reviewer for several usefulcomments. The research work at Physical Research Lab-oratory is funded by the Department of Space, Govern-ment of India. IZ is supported by the Russian Founda-tion for Basic Research (RFBR) No. 17-52-45020 andby the IAP RAS state program 0035-2014-0030. TB ac-knowledges funding from the National Natural ScienceFoundation of China through grant 11633005.
REFERENCESAnathpindika, S. V. 2010, MNRAS, 405, 1431Anderson, L. D., & Bania, T. M. 2009, ApJ, 690, 706Andr´e, P., Di Francesco, J., Ward-Thompson, D., et al. 2014, inProtostars and Planets VI, ed. H. Beuther et al. (Tucson, AZ;Univ. Arizona Press), 27Assirati, L., Silva, N. R., Berton, L., Lopes, A. A., & Bruno,O. M. 2014, Journal of Physics: Conference Series, 490(1), 2014Baug, T., Dewangan, L. K., Ojha, D. K.,& Ninan, J. P. 2016,ApJ, 833, 85Baug, T., Dewangan, L. K., Ojha, D. K., et al. 2018, ApJ, 852,119Benjamin, R. A.,Churchwell, E., Babler, B. L., et al. 2003, PASP,115, 953Bergin, E. & Tafalla, M. 2007, ARA&A, 45, 339Beuther, H., Schilke, P., Menten, K. M., et al. 2002, ApJ, 566, 945Beuther, H. & Steinacker, J. 2007, ApJ, 656, 85Beuther, H., Ragan, S. E., Johnston, K., et al. 2015, A&A, 584, 67Bisbas, T. G., W¨unsch, R., Whitworth, A. P., & Hubber, D. A.2009, A&A, 497, 649Bisbas, T. G., Tanaka, K. E. I., Tan, J. C., Wu, B., & Nakamura,F. 2017, arXiv:1706.07006 Bohlin, R. C., Savage, B. D., & Drake, J. F. 1978, ApJ, 224, 13233Bressert, E., Bastian, N., Gutermuth, R., et al. 2010, MNRAS,409, 54Carey, S. J., Noriega-Crespo, A., Price, S. D., et al. 2005, BAAS,37, 1252Clemens, D. P., Pavel, M. D., & Cashman, L. R. 2012, ApJS, 200,21Davis, L., Jr., & Greenstein, J. L. 1951, ApJ, 114, 206Dewangan, L. K., Luna, A., Ojha, D. K., et al. 2015, ApJ, 811, 79Dewangan, L. K., Ojha, D. K., Luna, A., et al. 2016, ApJ, 819, 66Dewangan, L. K., Ojha, D. K., & Zinchenko, I. 2017, ApJ, 851,140Dewangan, L. K. 2017, ApJ, 837, 44Dewangan, L. K., & Ojha, D. K. 2017, ApJ, 849, 65Dewangan, L. K., Devaraj, R., & Ojha, D. K. 2018, ApJ, 854, 106Evans, N. J., II, Dunham, M. M., Jørgensen, J. K., et al. 2009,ApJS, 181, 321Fallscheer, C., Beuther, H., Zhang, Q., Keto, E., & Sridharan,T. K. 2009, A&A, 504, 127Flaherty, K. M., Pipher, J. L., Megeath, S. T., et al. 2007, ApJ,663, 1069Fujita, S., Torii, K., Kuno, N., et al. 2017, arXiv:1711.01695Fukui, Y., Ohama, A., Hanaoka, N., et al. 2014, ApJ, 780, 36Fukui, Y., Torii, K., Ohama, A., et al. 2016, ApJ, 820, 26Fukui, Y., Kohno, M., Yokoyama, K., et al. 2018, PASJ, tmp, 53Furukawa, N., Dawson, J. R., Ohama, A., et al. 2009, ApJL, 696,L115Gonzalez, R, & Woods, R. 2011,
Digital Image Processing (Pearson Education) ISBN 9780133002324Gutermuth, R. A., Megeath, S. T., Myers, P. C., et al. 2009,ApJS, 184, 18Gutermuth, R. A., Heyer,M., 2015, AJ, 149, 64Habe, A., & Ohta, K. 1992, PASJ, 44, 203Hayashi, K., Sano, H., Enokiya, R., et al. 2018, PASJ, tmp, 10Haworth, T. J., Tasker, E. J., Fukui, Y., et al. 2015a, MNRAS,450, 10Haworth, T. J., Shima, K., Tasker, E. J., et al. 2015b, MNRAS,454, 1634Helfand, D. J., Becker, R. H., White, R. L., Fallon, A., & Tuttle,S. 2006, AJ, 131, 2525Inoue, T., & Fukui, Y. 2013, ApJL, 774, 31Jackson, J. M., Rathborne, J. M., Shah, R. Y., et al. 2006, ApJS,163, 145Kohno, M., Torii, K., Tachihara, K., et al. 2018, PASJ, tmp, 8Lawrence, A., Warren, S. J., Almaini, O., et al. 2007, MNRAS,379, 1599Lu, Xing, Zhang, Q., Liu, H. B., Wang, J., & Gu, Q. 2014, ApJ,790, 84Mallick, K. K., Ojha, D. K., Tamura, M., et al. 2015, MNRAS,447, 2307Matsakis, D. N., Evans, N. J., II, Sato, T., & Zuckerman, B.1976, AJ, 81, 172Molinari, S., Swinyard, B., Bally, J., et al. 2010, A&A, 518, L100Ohama, A., Dawson, J. R., Furukawa, N., et al. 2010, ApJ, 709,975Ohama, A., Kono, M., Fujita, S., et al. 2017a, arXiv:1706.05659Ohama, A., Kohno, M., Hasegawa, K., et al. 2017b,arXiv:1709.02320Oskinova, L. M., Todt, H., Ignace, R., et al. 2011, MNRAS, 416,1456Panagia, N. 1973, AJ, 78, 929Rathborne, J. M., Jackson, J. M., & Simon, R. 2006, ApJ, 641,389Rathborne, J. M., Simon, R., & Jackson, J. M. 2007, ApJ, 662,1082Rathborne, J. M., Jackson, J. M., Chambers, E. T., et al. 2010,ApJ, 715, 310Rebull, L. M., Guieu, S., Stauffer, J. R., et al. 2011, ApJS, 193, 25Sano, H., Enokiya, R., Hayashi, K., et al. 2018, PASJ, tmp, 26Schuller, F. et al. 2009, A&A, 504, 415Skrutskie, M. F. et al. 2006, AJ, 131, 1163Smith, L. J., Norris, R. P. F., & Crowther, P. A. 2002, MNRAS,337, 1309Sridharan, T. K., Beuther, H., Schilke, P., Menten, K. M., &Wyrowski, F. 2002, ApJ, 566, 931Tackenberg, J., Beuther, H., Henning, T., et al. 2014, A&A, 565,A101 loud-Cloud collision in IRAS 18223-1243 7 −20 −10 0 10 20+60+55+45+40+50
IRAS 18223−1243c) V e l o c it y ( k m / s ) Angular offset (arcmin)Sh 2−53
Figure 1. a) Herschel µ m image of a field (size ∼ ◦ .42 × ◦ .42) containing the sites IRAS 18223-1243 and Sh 2-53 in the directionof l = 17 ◦ .6 –19 ◦ . A solid box (in cyan) indicates the area shown in Figure 1b. b) The contours of the GRS CO emission are presentedwith levels of 72.2 K km s − × (0.22, 0.3, 0.35, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 0.98). The CO emission is integrated over a velocity from 37to 60 km s − . c) A p-v plot along the axis as shown in Figure 1b (see a dotted line in Figure 1b). In the first two top panels, a dashed boxshows the area studied by Baug et al. (2018), while a dotted-dashed box indicates the area investigated in this paper (see Figure 2a). Theposition of IRAS 18223-1243 is highlighted by a star in the panels “a” and “b”. L. K. Dewangan et al. fl1fl2fl1fl2
IRDC 18223
Figure 2. a) A false color
Herschel µ m image of IRAS 18223-1243 (area ∼ ◦ .4 × ◦ .4) is overlaid with the MAGPIS 20 cm continuumcontours. The contours (in magenta) are shown with levels of 1.5, 2.5, and 5 mJy/beam. Two filamentary features are highlighted byarrows (in white). b) The contours of the ATLASGAL 870 µ m continuum emission are presented with levels of 0.1, 0.2, 0.4, 1.0, 1.4, and2.0 Jy/beam. The dotted red box highlights the area shown in Figure 2c. c) Overlay of MAGPIS 20 cm continuum contours on the Spitzer µ m image. The 20 cm continuum contours (in red) are the same as in Figure 2a. A contour (in cyan) is also overlaid on the 8.0 µ mmap, tracing an extended emission seen in the 8 µ m image. In the direction of IRDC 18223 (length ∼ loud-Cloud collision in IRAS 18223-1243 9 −0.1−0.2 G a l ac ti c L a tit ud e [ d e g r ee ] Galactic Longitude [degree]
Figure 3.
The CO(J =1 −
0) velocity channel contour maps. The molecular emission is integrated over a velocity interval, which is givenin each panel (in km s − ). In each panel, the contours are shown with the levels of 1.4, 2.5, 3.5, 5, 6, 7.5, and 10 K km s − . The positionof IRAS 18223-1243 is highlighted by a star in each panel.Takahira, K., Tasker, E. J., & Habe, A. 2014, ApJ, 792, 63Takahira, K., Shima, K., Habe, A., & Tasker, E. J. 2018, PASJ,tmp, 42Tan, J. C., Beltr´an, M. T., Caselli, P., Fontani, F., Fuente, A.,Krumholz, M. R., McKee, C. F., Stolte, A. 2014, in Protostarsand Planets VI, ed. H. Beuther et al. (Tucson, AZ: Univ.Arizona Press), 149 Torii, K., Hattori, Y., Hasegawa, K., et al. 2017a, ApJ, 835, 142Torii, K., Matsuo, M., Fujita, S., et al. 2017b, arXiv:1710.08564Williams, J. P., de Geus, E. J., & Blitz, L. 1994, ApJ, 428, 693Yan, Q. Z., Xu, Y., Zhang, B., et al. 2016, AJ, 152, 117Zhang, C. P., Yuan, J. H., Li, G. X., Zhou, J. J., & Wang, J. J.2017, A&A, 598, A76Zinnecker, H., & Yorke, H. W. 2007, ARA&A, 45, 481 Angular offset V e l o c i t y ( k m / s ) Bridge featureIRASLocation b) Figure 4. a) Integrated CO (J=1-0) emission map of the region around IRAS 18223-1243. The CO integrated velocity range is from37 to 57 km s − . In the molecular map, the areas of six small fields (i.e. ar1 to ar6) are also highlighted by boxes. A solid line representsthe axis (length ∼
23 pc), where a p-v diagram is extracted in Figures 4b. A star symbol indicates the position of IRAS 18223-1243. b)A p-v diagram along the axis as shown in Figure 4a, tracing at least three cloud components (at ∼ ∼
51, and ∼
55 km s − ) along theline-of-sight. A relatively weak CO emission between two velocity peaks (at ∼
45 and ∼
51 km s − ) is highlighted by an arrow. A verticaldotted line indicates the position of IRAS 18223-1243. c to h panels) The GRS CO(1-0) spectra in the direction of six small fields (i.e.ar1 to ar6; see corresponding boxes in Figure 4a). Each spectrum is obtained by averaging each area. loud-Cloud collision in IRAS 18223-1243 11
Figure 5.
The CO emission integrated over four different velocity ranges (at 37–47, 47.5–49.5, 50–53.5, and 54–57 km s − ) is presented.The contour levels are 25, 30, 35, 40, 50, 60, 70, 80, 90, and 95% of the peak value (in K km s − ), which is given in each panel. Theposition of IRAS 18223-1243 is highlighted by a star in each panel. Figure 6. a) Integrated intensity map of CO (J = 1-0) from 37 to 57 km s − (see also Figure 4a). The contour levels are 25, 30,35, 40, 50, 60, 70, 80, 90, and 95% of the peak value (i.e. 61.75 K km s − ). b) Latitude-velocity map of CO. The CO emission isintegrated over the longitude from 18 ◦ .42 to 18 ◦ .82. c) The CO emission integrated over two different velocity ranges (at 37–47 and49–57 km s − ) is presented. The contour levels of the background CO emission map are 39.74 K km s − × (0.25, 0.3, 0.35, 0.4, 0.5,0.6, 0.7, 0.8, 0.9, and 0.95), while the CO contours (in blue) are 31.72 K km s − × (0.25, 0.3, 0.35, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, and0.95). d) Longitude-velocity map of CO. The CO emission is integrated over the latitude from − ◦ .3 to 0 ◦ .078. In each left panel (i.e.Figures 6a and 6c), the position of IRAS 18223-1243 is highlighted by a star. In Figures 6b and 6d, two velocity peaks are marked bybroken lines, and are interconnected by a relatively weak CO emission (see also Figure 4b). A dotted-dashed line indicates the positionof IRAS 18223-1243 in each right panel (i.e. Figures 6b and 6d). loud-Cloud collision in IRAS 18223-1243 13
Figure 7. a) The panel shows the resulting
Herschel temperature map of IRAS 18223-1243. b) The panel presents the resulting
Herschel column density ( N (H )) map of IRAS 18223-1243. One can also compute the extinction value with A V = 1 . × − N (H ) (Bohlinet al. 1978). c) The clumps identified using the Herschel column density map are shown by upside down triangles, and the extension ofeach
Herschel clump is also highlighted along with its corresponding clump ID (see also Table 2). In each panel, the position of IRAS18223-1243 is highlighted by a star. fl2fl1
Figure 8. a) A two color-composite map (red:
Herschel column density map; green:
Herschel µ m image) of IRAS 18223-1243. Here,the column density map is processed through an “Edge-DoG” algorithm. The 20 cm continuum contours (in black) are also overlaid onthe composite image, and are the same as in Figure 2a. b) The CO emission maps at 37–47 and 49–57 km s − . The maps are the sameas in Figure 6c. Two arrows indicate nearly overlapped zones of clouds. c) The CO emission filled contour at [37, 47] km s − is shownwith the level of 14 K km s − , while the CO emission contour (in red) at [49, 57] km s − is overlaid with the level of 12 K km s − . loud-Cloud collision in IRAS 18223-1243 15 Figure 9.
Selection of YSOs in the region around IRAS 18223-1243 (see Figure 2a). a)
Spitzer color-magnitude plot (i.e. [3.6] − [24] vs[3.6]) of sources. The plot helps us to depict YSOs belonging to different evolutionary stages (see dashed lines). The boundary of YSOsagainst contaminated candidates (galaxies and disk-less stars) is highlighted by dotted-dashed lines (in red) (see Rebull et al. 2011, formore details). Flat-spectrum and Class III sources are shown by “ (cid:51) ” and “ (cid:50) ” symbols, respectively. b) Spitzer color-color plot ([5.8] − [8.0]vs [3.6] − [4.5]) of sources. The PAH-emitting galaxies and the PAH-emission-contaminated apertures are marked by “*” and “ × ” symbols,respectively (see the text). Class III sources are shown by black squares in the plot. c) Spitzer color-color plot ([4.5] − [5.8] vs [3.6] − [4.5])of sources. d) NIR color-magnitude plot (H − K/K) of sources. In all the panels, Class I YSOs and Class II YSOs are represented by redcircles and open blue triangles, respectively. In the last three panels, the dots in gray color show the stars with only photospheric emissions.Considering the large numbers of stars with photospheric emissions, we show only some of these stars in the last three panels. In the firstthree panels, an extinction vector is plotted (e.g. Flaherty et al. 2007).
Figure 10.
Spatial distribution of YSOs in the region around IRAS 18223-1243. a) Overlay of all the identified YSOs on the
Herschel column density map. The YSOs (Class I (circles), Flat-spectrum (diamond), and Class II (triangles)) are overlaid on the
Herschel columndensity map. The YSOs identified using the
Spitzer color-magnitude scheme (i.e. [3.6] − [24] vs [3.6]; see Figure 9a), four Spitzer -GLIMPSE3.6-8.0 µ m bands (see Figure 9b), three Spitzer -GLIMPSE 4.5-8.0 µ m bands (see Figure 9c), and NIR color-magnitude scheme (i.e. H − K/K;see Figure 9d) are shown by magenta, red, cyan, and blue color symbols, respectively. b) Overlay of surface density contours of YSOs (inwhite) on the
Herschel column density map. c) Overlay of surface density contours of YSOs (in red) on the integrated intensity mapsof CO. The background molecular maps are the same as in Figure 6c. In the first two panels, the background map is the same as inFigure 7b. In the last two panels, the surface density contour levels are 2, 3, and 5 YSOs/pc . The Herschel clumps are also highlightedby green upside down triangles in the last two panels (see Figure 7c and also Table 2). loud-Cloud collision in IRAS 18223-1243 17
Figure 11.
A false color
Herschel µ m image is overlaid with the GPIPS H-band polarization vectors (in white) of stars. b) Overlayof mean GPIPS polarization vectors (in red) on the molecular map. The background molecular maps are the same as in Figure 6c. Areference vector of 5% is highlighted in each panel. Table 2
Herschel clumps and their physical parameters (see Figure 7c).ID l b R clump M clump (degree) (degree) (pc) ( M (cid:12)(cid:12)