Chemical Evolution of HC3N in Dense Molecular Clouds
aa r X i v : . [ a s t r o - ph . GA ] O c t Mon. Not. R. Astron. Soc. , 000–000 (0000) Printed 17 October 2019 (MN L A TEX style file v2.2)
Chemical Evolution of HC N in Dense Molecular Clouds
Naiping Yu , ⋆ , Jun-Jie Wang , ,Jin-Long Xu National Astronomical Observatories, Chinese Academy of Sciences, Beijing 100012, China NAOC-TU Joint Center for Astrophysics, Lhasa 850000, China
17 October 2019
ABSTRACT
We investigated the chemical evolution of HC N in six dense molecular clouds, using archivalavailable data from the Herschel infrared Galactic Plane Survey (Hi-GAL) and the MillimeterAstronomy Legacy Team Survey at 90 GHz (MALT90). Radio sky surveys of the Multi-Array Galactic Plane Imaging Survey (MAGPIS) and the Sydney University Molonglo SkySurvey (SUMSS) indicate these dense molecular clouds are associated with ultracompact HII(UCHII) regions and/or classical HII regions. We find that in dense molecular clouds as-sociated with normal classical HII regions, the abundance of HC N begins to decrease orreaches a plateau when the dust temperature gets hot. This implies UV photons could de-stroy the molecule of HC N. On the other hand, in the other dense molecular clouds as-sociated with UCHII regions, we find the abundance of HC N increases with dust temper-ature monotonously, implying HC N prefers to be formed in warm gas. We also find thatthe spectra of HC N (10-9) in G12.804-0.199 and RCW 97 show wing emissions, and theabundance of HC N in these two regions increases with its nonthermal velocity width, indi-cating HC N might be a shock origin species. We further investigated the evolutionary trendof N (N H + )/ N (HC N) column density ratio, and found this ratio could be used as a chemicalevolutionary indicator of cloud evolution after the massive star formation is started.
Key words: astrochemistry - stars: formation - ISM: clouds - ISM: abundance
Carbon-chain species account for a substantial fraction of the in-terstellar molecules observed so far. They are prone to be de-pleted onto dust grains when the gas is cold, and destroyed byUV radiations (Sakai & Yamamoto 2013). In star-forming regions,many carbon-chain species could be used as “chemical clocks” totrace star formation (e.g. Suzuki et al. 1992; Hirota et al. 2009).Cyanopolyynes (HC n + N) are one of the representative carbon-chain species. Since the first detection of interstellar cyanoacety-lene (HC N) by Turner (1971) in Sgr B2, cyanopolyynes have beenfound to be ubiquitously in our Galactic interstellar medium (ISM)(e.g., Cernicharo & Gu´ e lin 1996; Takano et al. 1998; Crovisier etal. 2004). Previously, long chain cyanopolyynes were believed to beabundant in cold dark clouds. In hot cores, they could not be formedefficiently (Millar 1997). However, long chain cyanopolyynes ofHC N, HC N and HC N were detected by Sakai et al. (2008) in theprotostar IRAS 04368+2557L1527. They proposed a new chem-istry called “warm carbon-chain chemistry (WCCC)” in a warmand dense region near the low-mass protostars. Hassell et al. (2008)then made a chemical model of this region. Their calculations showthe cyanopolyynes abundance enrichment in the gas phase as thegrains warm up to 30 K. Chapman et al. (2009) further presented ⋆ E-mail: [email protected] that cyanopolyynes could be formed under a hot core condition andshow as “chemical clocks” to determine the age of hot cores.HC N is the simplest form of cyanopolyynes. This moleculetraces both dense and warm gas. In warm gas, it could be formedfrom CH (Hassel et al. 2008) and/or C H (Chapman et al. 2009)evaporated from grain mantles. Sanhueza et al. (2012) found themedian value of HC N column density increases as a function ofclump evolutionary stage. Taniguchi et al. (2016) observed three C isotopologues of HC N in L1527 and G28.28-0.36. They foundthe abundance of H CCCN and HC CCN are comparable, whileHCC CN is more abundant. This result could be explained by thatHC N might be formed from the neutral-neutral reaction betweenC H and CN: C H + CN −→ HC N + H. The abundance of HC Ncan also be enhanced after the passage of shocks. Mendoza et al.(2018) found the abundance of HC N increases by a factor of 30 inthe shocked region of L1157. Taniguchi et al. (2018) carried out ob-servations of HC N and HC N toward 52 high-mass star-formingregions with the Nobeyama 45 m telescope. They found the spec-tra of some HC N show wing emissions, suggesting HC N is anoutflow shock origin species.The destruction of HC N could be caused by UV radiations.Yu & Xu (2016) found the fractional abundances of HC N de-crease as a function of Lyman continuum fluxes in a number ofRed MSX (Midcourse Space Experiment) Sources (RMSs), indi-cating this molecule could be destroyed by UV photons. Urquhartet al. (2019) conducted a 3-mm molecular line survey towards 570 © 0000 RAS
TLASGAL (APEX Telescope Large Area Survey of the Galaxy)clumps. They found the detection rate of HC N (10-9) increasesfrom the “Quiescent” stage to the “Protostellar” stage, and reachesa plateau in the “Young Stellar Object (YSOs)” and “HII region”stages. They guess that in the late two stages the formation of HC Nis in equilibrium with its destruction by UV photons or other chem-ical reactions. Even to today, there are few researches about chem-ical evolution of HC N in massive star-forming regions. Previousstudies mentioned above are surveys of massive clumps/cores indifferent giant molecular clouds (GMCs). The distances and initialconditions may be quite different in different GMCs. For example,the measured C/ C ratio ranges from ∼
20 to ∼
70, dependingon the distance to the Galactic center (e.g. Savage et al. 2002). Thismight complicate quantitative comparisons and make statistical re-sults not significant. In our previous paper (Yu et al. 2018 here-after YXW18), we studied the chemical evolution of N H + usingdata from MALT90 and Hi-Gal in six massive star-forming regions.Here we present our study of HC N instead of N H + . The study ofother molecules such as C H, HCO + and HNC will come in anotherpaper. The distances and initial conditions of clumps could be re-garded as the same in the same cloud, and thus a comparison ofthe chemical evolution in different clumps along the evolutionarysequence is valid. We introduce our data in Section 2, results anddiscussions are in Section 3, and finally we summarize in Section4. Our molecular line data of HC N (10-9) comes from MALT90, andthe dust infrared data is from Hi-GAL. As the method described byYXW18, we first calculate the H column density and dust tem-perature maps of these regions, and then the column density andabundance maps of HC N. Here we give a brief introduction.The Hi-GAL data set is comprised of 5 continuum images ofthe Milky Way Galaxy using the PACS (70 and 160 µ m) and SPIRE(250, 350 and 500 µ m) instruments. Following the steps describedby Wang et al. (2015), we made H column density and dust tem-perature maps of each region through the method of spectral energydistribution (SED). After removing the background and foregroundemissions, we convolved all the images of Hi-GAL to a spatial res-olution of 45 ′′ , which is the measured beamsize of Hi-GAL obser-vations at 500 µ m (Traficante et al. 2011). For each pixel, we useequation I ν = B ν (1 − e − τ ν ) (1)to model intensities at various wavelengths. The optical depth τ ν could be estimated through τ ν = µ H m H κ ν N H / R gd (2)We adopt a mean molecular weight per H molecule of µ H = 2.8 toinclude the contributions from Helium and other heavy elements. m H is the mass of a hydrogen atom. N H is the column density. R gd is the gas-to-dust mass ratio which is set to be 100. According toOssenkopf & Henning (1994), dust opacity per unit dust mass ( κ ν )could be expressed as κ ν = . ν GHz ) β cm g − (3)where the value of the dust emissivity index β is fixed to 1.75 in ourfitting. The two free parameters ( N H and T d ) for each pixel couldbe fitted finally. The final resulting dust temperature and column density maps, which have a spatial resolution of 45 ′′ with a pixelsize of 13 ′′ , are shown in Figures 1-6.MALT90 is an international project with the aim to character-ize physical and chemical properties of massive star formation inour Galaxy (e.g., Foster et al. 2011; Foster et al. 2013; Jackson etal. 2013). This project was carried out with the Mopra Spectrom-eter (MOPS) arrayed on the Mopra 22 m telescope. The beamsizeof Mopra is 38 ′′ at 86GHz, with a beam efficiency between 0.49 at86 GHz and 0.42 at 115 GHz (Ladd et al. 2005). The target of thissurvey are selected from the ATLASGAL clumps found by Con-treras et al. (2013). The image size of each MALT90 data cube isabout 4 ′′ × ′′ , with a step of 9 ′′ . We downloaded the data filesfrom the MALT90 Home Page , and assembled all the MALT90data into a new data cube in a certain region if they have the samevelocity component. We have found out six dense molecular cloudsshowing distinct emissions of HC N (10-9). Their infrared imagesand new combined integrated emissions of HC N (10-9) are alsoshown in Figures 1-6 respectively. All sources involve at least twoATLASGAL clumps. To calculate the abundance of HC N in eachpixel, we also smoothed the molecular data into a new beamsizeof 45 ′′ with a new step of 13 ′′ . By assuming local thermodynamicequilibrium (LTE) conditions and HC N (10-9) is optically thin, wecalculated the HC N column density in each pixel where its emis-sion is greater than 5 σ , using the equation from Sanhueza et al.(2012): N = πν c Q rot g u A ul exp ( E l / kT ex )1 − exp ( − h ν/ kT ex ) R T mb dvJ ( T ex ) − J ( T bg ) (4)where c is the velocity of light in the vacuum, g u is the statisti-cal weight of the upper level, A ul is the Einstein coefficient forspontaneous transition, E l is the energy of the lower level, Q rot is the partition function, T bg is the background temperature, T ex is the excitation temperature. Like the assumption make by San-hueza et al. (2012), we here also assume that T ex is equal to thedust temperature derived above. The values of g u , A ul and E l couldbe found in the Cologne Database for Molecular Spectroscopy(CDMS) (M¨uller et al. 2001, 2005 ). J ( T ) is defined by J ( T ) = h ν k e h ν/ kT − N, we made an assumptionthat the HC N (10-9) line is optically thin. The true column den-sity derived by Eq. (4) should be multiplied by a factor of τ / (1- e − τ ). For an intermediately optically thick line ( τ : 0.5 ∼ N ( χ (HC N)) for each pixelcan be calculated through χ (HC N) = N (HC N)/ N (H ) finally. TheHC N abundance maps for each source are shown in Figures 7-12.
The dense cloud G5.899-0.429 involves 5 ATLASGAL clumps.Four of them have been observed by MALT90 (Figure 1, the greenboxes). The distance of this cloud is about 2.9 kpc (Sato et al. http://atoa.atnf.csiro.au/MALT90 N (10-9) emis-sion is very compact and comes from the densest part of thiscloud. The most massive clump AGAL005.884-00.392, which hasthe strongest emission of HC N, is also known as an expandingUCHII region W28 A2 (Wood & Churchwell 1989). The expand-ing velocity of this UCHII regions is about 35 km s − (Acord et al.1998). Near infrared observations indicate the exciting source ofthis UCHII region is a young O-type star (Feldt et al. 2003). Zapataet al. (2019) carried out high angular resolution observations andfound an explosive outflow from this UCHII region. The south-eastpart of this cloud involves two ATLASGAL clumps which showrelatively weak emissions of HC N (10-9). The 90 cm radio con-tinuum emissions from MAGPIS are shown in yellow contours inthe top left panel of Figure 1. We can see that radio emissions hereare more diffuse and larger than that in the UCHII region, indi-cating this might be a normal classical HII region. From Figure7, we can see that compared to the south-east part, χ (HC N) ismore abundant in the UCHII region, and the abundance of HC Nincreases with dust temperature monotonously in the whole region.The spectra of HCO + (1-0) shows the so-called “blue profile” withextended wing emissions where the χ (HC N) is highest, indicatingstar-forming activities such as infall and outflow. This result sug-gests that HC N prefers to be formed in warm gas with massivestar-forming activities.
This dense molecular cloud is associated with the well-knownGMC W33. The distance of W33 is about 2.4 kpc (Immer et al.2013). It includes three large dust clumps (W33 Main, W33 Aand W33 B) and three smaller clumps (W33 Main1, W33 A1 andW33 B1). Even though these clumps are involved in a whole star-forming complex, radio line observations found W33 Main andW33 A have a radial velocity of ∼
36 km s − , while W33 B has adifferent radial velocity of ∼
58 km s − . Using molecular line datafrom MALT90, we checked both the two velocity components, onlyfounding HC N (10-9) emissions in W33 Main and W33 A (see thetop right panel of Figure 2). The 90 cm radio continuum emissionsfrom MAGPIS in Figure 2 show W33 Main as a compact source,which is also known to be an UCHII region (Keto & Ho 1989).On the south-east of W33 Main, there is a strong arc-shaped ra-dio emission. Ho et al. (1986) suggest this is an ionization frontpenetrating W33 Main. The HCO + (1-0) spectra in W33 Main alsoshows the so-called “blue profile” with extended wing emissions(Figure 8), indicating infall and outflow activities in W33 Main. Wedo not found radio emission in the center of W33 A. However, vander Tak & Menten (2005) found faint 43 GHz radio emissions inW33 A with higher resolution observations. They suggest the faintemissions come from an ionised wind or a hypercompact HII region(HCHII) in W33 A. Immer et al. (2014) detected a large numberof simple and complex molecules in W33 A. They suggest W33 Amay be in the transition from the hot core stage to the HCHII regionphase. Thus W33 Main is more evolved than W33 A. From Figure8, we can see that HC N is more abundant in W33 Main than that inW33 A. Like that found in G5.899-0.429, the abundance of HC Nalso increases with dust temperature in G12.804-0.199. This resultalso suggests that HC N prefers to be formed in warm gas withmassive star-forming activities.
The dense molecular cloud G326.641+0.612 is associated with theclassical HII region RCW 95 (Rodgers et al. 1960). The kinematicdistance of RCW 95 is about 2.4 kpc (Giveon et al. 2002). YXW18found that in this region the abundance of N H + reaches a plateauas the dust temperature is above 27 K (see their Figure 10). Theythus suggest the destruction of N H + by CO or UV photons aroundthis classical HII region. From Figure 9, we can see that as the dusttemperature gets hot, the abundance of HC N also seems to reacha plateau, indicating the destruction of HC N by UV photons.
The dense molecular cloud G327.293-0.579 is associated with aluminous photon dominated region (PDR) around the classical HIIregion RCW 97 on the north side, and an IRDC on the south side(Wyrowski et al. 2006). From Figure 4, we can see that the emis-sion of HC N (10-9) mainly comes from the IRDC which hosts thehot core G327.3-0.6 (Gibbet al. 2000) and extended green object(EGO) candidate G327.30-0.58 (Cyganowski et al. 2008). Assum-ing a kinematic distance of 3.0 kpc (Russeil 2003), the Lyman con-tinuum flux from RCW97 will be more than 10 photons s − (Conti& Crowther 2004), indicating this is a giant HII region. Both infalland outflow activities have been found by Leurini et al. (2017) inthe IRDC. From Figure 10, it can be noticed that the abundanceof HC N is highest in the IRDC, and begins to drop as the dusttemperature gets hotter than 30 K. This is also consistent with thescenario that HC N could be destroyed by UV photons.
This dense molecular cloud is associated with the infrared bubbleS36 (Churchwell et al. 2006). The radio continuum emissions fromSUMSS shown in Figure 5 indicate this is also a classical HII re-gion. YXW18 found that in this region, the abundance of N H + increases with dust temperature when it is below 28 K, and thendecreases quickly in the PDR where T d is hotter than 28 K (alsosee their Figure 10). From Figure 11, we can see that the situationof HC N is similar to that of N H + in this region, indicating UVphotons are also destroying HC N and N H + on the PDR of S36. The dense molecular cloud G345.448+0.314 involves two ATLAS-GAL clumps. The two clumps are associated with IRAS 17008-4040 and IRAS 17009-4042 respectively. The radio continuumemissions from SUMSS in Figure 6 indicate HII regions in thesetwo clumps. Using high resolution archival data from the Giant Me-trewave Radio Telescope (GMRT), Dewangan et al. (2018) found13 HII regions with radius in the range of 0.06 pc and 0.25 pc inthese two IRAS sources. The radius of these HII regions indicatethey are still in the UCHII stage. Like those found in G5.899-0.429and G12.804-0.199 introduced above, the abundance of HC N alsoincreases with dust temperature (Figure 12), indicating the produc-tion of HC N is more efficient than its destruction here. This maybe because compared to classical HII regions, UCHII regions arestill surrounded by dense gas, providing shielding against UV radi-ation. The spectra of HCO + (1-0) shows red and blue profiles withwing emissions in IRAS 17008-4040 and IRAS 17009-4042, in-dicating star-forming activities in this dense molecular cloud. Like3hat found in the other UCHII regions (G5.899-0.429 and G12.804-0.199) above, this result also suggests that HC N prefers to beformed in warm gas with massive star-forming activities.
We found that in dense molecular clouds associated with classi-cal HII regions (RCW 95, RCW 97 and infrared bubble S36),the abundance of HC N does not increase with dust temperaturemonotonously. It begins to decrease or reaches a plateau as the dusttemperature gets hot. In a previous paper (Yu & Xu 2016), we alsofound that the abundance of HC N decreases with Lyman contin-uum flux. These studies indicate that HC N can be destroyed by UVradiation. Chemical network from KIDA (Wakelam et al. 2014)tells us that HC N could be destroyed through reaction:HC N +Photon −→ CN + C H and/or HC N + Photon −→ HC N + + e − .Besides, UV photons could also destroy C H which is the mainprogenitor of HC N through reaction: C H + Photon −→ C H +H, leading to the production of HC N ineffective.On the other hand, the situation was quite different in UCHIIregions of G5.899-0.429, G12.804-0.199 and G345.448+0.14. Wefound that in these regions, the abundance of HC N increases withdust temperature. This may be because that in warm gas the pro-genitors of HC N (such as C H and CH ) could be easily releasedinto gas phase. Yu & Wang (2015) found that in massive youngstellar objects (MYSOs), the line widths of HC N are compara-ble to those of N H + , which is regarded as a good tracer of colddense gas. However, in UCHII regions the line widths of HC Nbecome broader than those of N H + . Taniguchi et al. (2018) alsofound that the line widths of HC N are significantly broader thanthose of HC N. These studies indicate HC N prefers to exist inmore active star-forming regions. From Figure 7, 8, 10 and 12, wecan see that the spectra of HCO + (1-0) show red and blue profileswith wing emissions where the abundance of HC N is highest. Pre-vious multi-wavelength observations also indicate shock activities(caused by infall, outflow and/or expanding HII regions) in theseregions. We also checked the spectra of HC N (10-9) in these re-gions, and found that in G12.804-0.199 and RCW 97 the spectraof HC N (10-9) show wing emissions (Figure 13). This suggestsHC N might be an outflow shock origin species. The abundance ofHC N could be increased in the passage of shocks. The chemicalmodel of Mendoza et al. (2018) shows that the abundance of HC Ncould be directly increased due to mantle sputtering due to the pas-sage of shocks. Besides, their model also indicates shock activitiescould increase the reaction efficiency of CN with C H : C H + CN −→ HC N + H.The velocity dispersion of HC N (10-9) caused by thermalmotions could be estimated by: ∆ V therm = s ln kT d ( 1 m HC N + m ) (6)where m HC N is the mass of HC N (51 per amu), m is the meanmolecular mass (2.3 per amu). Figure 13 shows the relationshipbetween the abundance of HC N and its nonthermal line widths ( ∆ V nontherm ≡ ∆ V FWHM - ∆ V therm ) in G12.804-0.199 and RCW 97. Itcan be noticed that the abundance of HC N increases with its non-thermal line width in these two regions. This result also suggeststhat HC N could be efficiently formed by massive star formation http://kida.obs.u-bordeaux1.fr/ activities. We thus regard that HC N prefers to be formed in warmgas with massive star-forming activities. We suggest more line ob-servations with higher resolutions to be carried out to found out thechemical evolution of HC N in massive star-forming regions.Previous studies of low-mass star-forming regions suggest theratio of N (nitrogen-bearing species)/ N (carbon-chain species) in-creases as a cloud evolves (e.g. Suzuki et al. 1992; Benson et al.1998; Hirota et al. 2009). Suzuki et al. (1992) carried out observa-tions of CCS, HC N, HC N and NH toward 49 dark cloud coresin the Taurus and Ophiuchus regions. They found carbon-chainmolecules like CCS are abundant in the early stages of chemicalevolution, while NH is abundant in the later stages. They supposethat in the early stages of star formation, carbon-chain moleculescould be efficiently formed from ionic carbon (C + ) and atomic car-bon (C). However, as the cloud evolves further, the formation ef-ficiency of carbon-chain molecules decrease, because most of thecarbon atoms are converted into the form of CO, which is chem-ically stable. On the other hand, nitrogen-bearing species such asNH become gradually abundant in the central part of the core.Thus the abundance ratio of CCS and NH could be used as agood indicator of cloud evolution and star formation. Benson etal. (1998) made a high spatial resolution observation of N H + ,C H and CCS toward 60 dense cores. They found that the ratioof N (CCS)/ N (N H + ) in starless cores is higher by a factor of 2 thanthat in cores with stars. This result is consistent with the finding ofSuzuki et al. (1992). Ohashi et al. (2014; 2016) have investigatedmolecular lines of HC N (10-9) and N H + (1-0) in the cluster form-ing regions of Orion A and Vela C GCMs. They found that the ra-tios of N (NH )/ N (CCS) and N (N H + )/ N (HC N) could also be thetracers of the chemical evolution even in the high mass star form-ing region in the same way as the low mass star forming region.Figure 14 shows the relationship between N (N H + )/ N (HC N) and T dust in all of our sources. It can be noticed that in all sources, therelative column density ratios of N H + and HC N do not increasewith dust temperature. Moreover, in G5.899-0.429, G12.804-0.199,RCW 97, S36 and G345.448+0.314, it is clear that this ratio de-creases as dust temperature increases, which is totally different tothat found by Ohashi et al. (2016). The reason may be that we fo-cused on the different evolutionary stages of massive star forma-tion. It is generally accepted that massive stars evolve from star-less cores in IRDCs to hot cores with central young stellar ob-jects, then to HCHII and UCHII regions. The final stages are com-pact and classical HII regions (Zinnecker et al. 2007). Previousstudies of Ohashi et al. (2016) compared starless cores with star-forming (Class 0/I) sources. In this work, we focused on sourcesof UCHII regions and classical HII regions, where massive pro-tostars have already formed. According to the new chemistry ofWCCC, Sakai et al. (2008) suggested that the HC N will be moreenhanced after star formation starts due to the evaporation fromthe grain surface. The increase of χ (HC N) with dust tempera-ture has been founded in G5.899-0.429 (Figure 7), G12.804-0.199(Figure 8) and G345.448+0.314 (Figure 12), and the decrease of χ (N H + ) with dust temperature in S36 has also been shown in ourprevious paper of YXW18. This may be the reason that the ratioof N (N H + )/ N (HC N) decreases with the dust temperature in oursources. Our study suggests this ratio still could be used as a chem-ical evolutionary indicator of cloud evolution after the massive starformation is started.4
SUMMARY
We investigate the chemical evolution of HC N in six dense molec-ular clouds, using data from MALT90 and Hi-GAL. Radio skysurveys indicate these dense molecular clouds are associated withUCHII regions and/or classical HII regions. We found that in densemolecular clouds associated with classical HII regions, the abun-dance of HC N decreases or reaches a plateau when the dust tem-perature gets hot, implying UV photons could destroy the moleculeof HC N. On the other hand, in dense molecular clouds associatedwith UCHII regions, we found the abundance of HC N increaseswith dust temperature monotonously. The spectra of HCO + (1-0)and HC N (10-9) in some sources show wing emissions. We alsofound that the abundance of HC N increases with its nonthermalvelocity width in G12.804-0.199 and RCW 97. These results sug-gest HC N prefers to be formed in warm gas with star-forming ac-tivities, and could be destroyed by UV photons in the late stagesof massive star formation. We also found that in most sources, thecolumn density ratio of N H + and HC N decreases with the dusttemperature. Our study seems to support that the column densityratio of N H + and HC N could still be used as a chemical evolu-tionary indicator of cloud evolution after the massive star formationis started.
We thank the anonymous referee for constructive suggestions.This paper has made use of information from the ATLASGALDatabase Server . The ATLASGAL project is a collaboration be-tween the Max-Planck-Gesellschaft, the European Southern Obser-vatory (ESO) and the Universidad de Chile. This research made useof data products from the Millimetre Astronomy Legacy Team 90GHz (MALT90) survey. The Mopra telescope is part of the Aus-tralia Telescope and is funded by the Commonwealth of Australiafor operation as National Facility managed by CSIRO. This paper issupported by the Youth Innovation Promotion Association of CAS. REFERENCES
Acord J. M., Churchwell E., Wood D. O. S. 1998, ApJ, 495, L107Benson P. J., Caselli P., & Myers P. C. 1998, ApJ, 506, 743Cernicharo J., Gu´ e lin M. 1996, A&A, 309, L27Chapman J. F., Millar T. J., Wardle M., Burton M. G., Walsh A. J.2009, MNRAS, 394, 221Churchwell, E., Povich, M. S., Allen, D., et al. 2006, ApJ, 649,759Conti P. S., Crowther P. A. 2004, MNRAS, 355, 899Contreras Y., Schuller F., Urquhart J. S., et al. 2013, A&A, 549,A45Crovisier J. et al. 2004, A&A, 418, 1141Cyrowski C. J., Whitney B. A., Holden E. et al. 2008, AJ, 136,2391Dewangan L. K., Baug T., Ojha D. K., Ghosh S. K. 2018, ApJ,869, 30Feldt M., Puga, E., Lenzen, R. et al. 2003, ApJ, 599, L91Foster J. B., Jackson J. M., Barnes P. J. et al. 2011, ApJS, 197, 25Gibb E., Nummelin A., Irvine W. M., Whittet D. C. B., BergmanP. 2000, ApJ, 545, 309 http://atlasgal.mpifr-bonn.mpg.de/cgi-bin/ATLASGAL − DATABASE.cgi
Giveon A., Sternberg A., Lutz D., Fruchtgruber H., Pauldrach, A.W. A. 2002, ApJ, 566, 880Hassel G. E., Herbst E., Garrod R. T. 2008, ApJ, 681, 1385Hirota T. Ohishi M. & Yamamoto S. 2009, ApJ, 699, 585Ho P. T. P., Klein R. I., Haschick A. D. 1986, ApJ, 305, 714Immer K., Reid M. J., Menten K. M., Brunthaler A., Dame T. M.2013, A&A, 553, A117Immer K., Galv´ a n-Madrid R., K¨ o nig C., Liu H. B., Menten K. M.2014, A&A, 572, A63Jackson, J. M., Rathborne, J. M., Foster, J. B., et al. 2013, PASA,30, 57Keto E. R., & Ho P. T. P. 1989, ApJ, 347, 349Ladd, N., Purcell, C., Wong, T., & Robertson, S. 2005, PASA, 22,62Lada C. J., Lombardi M., & Alves J. F. 2010, ApJ, 724, 687Leurini S., Herpin F., van der Tak F., Wrowski F., Herczeg G. J.,van Dishoeck E. F. 2017, A&A, 602, A70Mendoza E., Lefloch B., Ceccarelli C. et al. 2018, MNRAS, 475,5501Millar T. J., Macdonald G. H., Gibb A. G. 1997, A&A, 325, 1163M¨uller H. S. P., Thorwirth S., Roth D. A., & Winnewisser G. 2001,A&A, 370, L49M¨uller H. S. P., Schl¨oder F., Stutzki J., & Winnewisser G. 2005,J. Molec. Struct., 742, 215Ohashi S., Tatematsu K., Choi M., et al. 2014, PASJ, 66, 119Ohashi S., Tatematsu K., Fujii K., et al. 2016, PASJ, 68, 3Ossenkopf V., Henning T., 1994, A&A, 291, 943Rathborne J. M., Whitaker J. S., Jackson J. M., et al. 2016, Publ.Astron. Soc. Aust., 33, e030Rodgers A. W., Campbell C. T., & Whiteoak J. B. 1960, MNRAS,121, 103Russeil D. 2003, A&A, 397, 133Sakai N. Sakai T. Hirota T. Yamamoto S. 2008 ApJ, 672, 371Sakai N., & Yamamoto S. 2013, ChRv, 113, 8981Sanhueza, P., Jackson, J. M., Foster, J. B., Garay, G., Silva, A.,Finn, S. C., 2012, ApJ, 756, 60Sato M., Wu Y. W., Immer K., et al. 2014, ApJ, 793, 72Suzuki H., Yamamoto S., Ohishi M. et al. 1992, ApJ, 392, 551Taniguchi K., Saito M., Ozeki H. 2016, ApJ, 830, 106Taniguchi K., Saito M., Sridharan T. K., Minamidani T. 2018,ApJ, 854, 133Taniguchi K., Saito M., Sridharan T. K., Minamidani T. 2019,ApJ, 872, 154Takano S., et al. 1998, A&A, 329, 1156Traficante, A., Calzoletti, L., Veneziani, M., et al. 2011, MNRAS,416, 2932Turner B. E. 1971, ApJ, 163, L35Urquhart J .S., Figura C., Wyrowski F., et al. 2019, MNRAS, 484,4444van der Tak F. F. S., & Menten K. M. 2005, A&A, 437, 947Wakelam V., Vastel C., Aikawa Y., et al. 2014, MNRAS, 445,2854Wang, K., Testi, L., Ginsburg, A., et al. 2015, MNRAS, 450, 4043Wood D. O., Churchwell E. 1989, ApJS, 69, 831Wyrowski F., Menten K. M., Schilke P., et al. 2006, A&A, 454,L91Yamamoto T., Nakagawa N., & Fukui Y. 1983, A&A, 122, 171Yu N. P., & Wang J. J. 2015, MNRAS, 451, 2507Yu N. P., & Xu J. L. 2016, ApJ, 833, 248Yu N. P., Xu J. L., Wang J. J., Liu X. L. 2018, ApJ, 865, 135Zapata L. A., Ho P. T. P., Guzm´ a n E. et al. 2019,arXiv:1904.043855innecker H., Yorke H. W., 2007, ARA&A, 45, 481 6 .950 5.900 5.850-0.350-0.400-0.450 Galactic longitude G a l a c t i c l a t i t u d e G5.899-0.429
Figure 1.
Top left: Three colour mid-infrared image of G5.899-0.429 created using the Spitzer IRAC band filters (8.0 µ m in red, 4.5 µ m in green and 3.6 µ min blue). The green boxes indicate the four observed regions by MALT90. Top right: The new combined image of HC N from MALT90 data set. The emissionhas been integrated from 3 to 15 km s − . The unit of the color bar on the right is in K km/s. Bottom left: The H column density of this region built throughSED fitting. The unit of color bar is in cm − . Bottom right: The dust temperature map in color scale derived from the SED fitting. The unit of color bar is inK. The ATLASGAL 870 µ m emissions (in white) are superimposed with levels 0.24, 0.48, 0.96, 1.92 and 3.84 Jy/beam in each panel. The red pluses mark thecenter locations of ATLASGAL clumps. The yellow star marks the UCHII region W28 A2. The 90 cm radio continuum emissions (in yellow) from MAGPISare superimposed with levels 0.1, 0.2, 0.3, 0.4, and 0.5 Jy beam − . G a l a c t i c l a t i t u d e G12.804-0.199
Figure 2.
Top left: Three colour mid-infrared image of G12.804-0.199 created using the Spitzer IRAC band filters (8.0 µ m in red, 4.5 µ m in green and 3.6 µ m in blue). The green boxes indicate the observed regions by MALT90. Top right: The new combined image of HC N from MALT90 data set. The emissionhas been integrated from 30 to 40 km s − . The unit of the color bar on the right is in K km/s. Bottom left: The H column density of this region built throughSED fitting. The unit of color bar is in cm − . Bottom right: The dust temperature map in color scale derived from the SED fitting. The unit of color bar is in K.The ATLASGAL 870 µ m emissions (in white) are superimposed with levels 0.24, 0.48, 0.96, 1.92, 3.84 and 7.68 Jy/beam in each panel. The red pluses markthe center locations of ATLASGAL clumps. The yellow stars mark W33A, W33A1, W33Main1, W33Main, W33B1 and W33B. The 90 cm radio continuumemissions (in yellow) from MAGPIS are superimposed with levels 0.02, 0.04, 0.08, 0.16, and 0.32 Jy beam − . G a l a c t i c l a t i t u d e G326.641+0.612
Figure 3.
Top left: Three colour mid-infrared image of G326.641+0.612 created using the Spitzer IRAC band filters (8.0 µ m in red, 4.5 µ m in green and 3.6 µ m in blue). The green boxes indicate the four observed regions by MALT90. Top right: The new combined image of HC N from MALT90 data set. Theemission has been integrated from -44 to -36 km s − . The unit of the color bar on the right is in K km/s. Bottom left: The H column density of this regionbuilt through SED fitting. The unit of color bar is in cm − . Bottom right: The dust temperature map in color scale derived from the SED fitting. The unit ofcolor bar is in K. The ATLASGAL 870 µ m emissions (in white) are superimposed with levels 0.42, 0.84, 1.68, 3.36, and 6.72 Jy/beam in each panel. The redpluses mark the center locations of ATLASGAL clumps. The 843 MHz SUMSS radio continuum emissions (in yellow) are superimposed with levels 0.2, 0.4,0.8, 1.6, and 3.2 Jy beam − . G a l a c t i c l a t i t u d e G327.293-0.579
Figure 4.
Top left: Three colour mid-infrared image of G327.293-0.579 created using the Spitzer IRAC band filters (8.0 µ m in red, 4.5 µ m in green and 3.6 µ min blue). The green boxes indicate the five observed regions by MALT90. Top right: The new combined image of HC N from MALT90 data set. The emissionhas been integrated from -50 to -41 km s − . The unit of the color bar on the right is in K km/s. Bottom left: The H column density of this regions built throughSED fitting. The unit of color bar is in cm − . Bottom right: The dust temperature map in color scale derived from the SED fitting. The unit of color bar is in K.The ATLASGAL 870 µ m emissions (in white) are superimposed with levels 0.21, 0.42, 0.84, 1.68, 3.36, and 6.72 Jy/beam in each panel. The red pluses markthe center locations of ATLASGAL clumps. The 843 MHz SUMSS radio continuum emissions (in yellow) are superimposed with levels 0.3, 0.6, 1.2, 2.4, and4.8 Jy beam − . G a l a c t i c l a t i t u d e G337.916-0.477
Figure 5.
Top left: Three colour mid-infrared image of G337.916-0.477 created using the Spitzer IRAC band filters (8.0 µ m in red, 4.5 µ m in green and 3.6 µ min blue). The green boxes indicate the six observed regions by MALT90. Top right: The new combined image of HC N from MALT90 data set. The emissionhas been integrated from -42 to -37 km s − . The unit of the color bar on the right is in K km/s. Bottom left: The H column density of this regions built throughSED fitting. The unit of color bar is in cm − . Bottom right: The dust temperature map in color scale derived from the SED fitting. The unit of color bar is in K.The ATLASGAL 870 µ m emissions (in white) are superimposed with levels 0.27, 0.54, 1.08, 2.16, and 4.32 Jy/beam in each panel. The red pluses mark thecenter locations of ATLASGAL clumps. The 843 MHz SUMSS radio continuum emissions (in yellow) are superimposed with levels 0.05, 0.10, 0.20, 0.40,0.80, 1.60 and 3.20 Jy beam − . G a l a c t i c l a t i t u d e G345.448+0.314
Figure 6.
Top left: Three colour mid-infrared image of G345.448+0.314 created using the Spitzer IRAC band filters (8.0 µ m in red, 4.5 µ m in green and 3.6 µ m in blue). The green boxes indicate the two observed regions by MALT90. Top right: The new combined image of HC N from MALT90 data set. Theemission has been integrated from -21 to -14 km s − . The unit of the color bar on the right is in K km/s. Bottom left: The H column density of this regionsbuilt through SED fitting. The unit of color bar is in cm − . Bottom right: The dust temperature map in color scale derived from the SED fitting. The unit ofcolor bar is in K. The ATLASGAL 870 µ m emissions (in white) are superimposed with levels 0.18, 0.36, 0.72, 1.44, and 2.88 Jy/beam in each panel. The redpluses mark the center locations of ATLASGAL clumps. The two yellow stars mark IRAS 17008-4040 and IRAS 17009-4042. The 843 MHz SUMSS radiocontinuum emissions (in yellow) are superimposed with levels 0.02, 0.04, 0.08, and 0.16 Jy beam − . dust (K) X ( HC N ) x - Figure 7.
Top: The HCO + (1-0) spectra superimposed on the calculated HC N abundance map of G5.899-0.429. Bottom: Abundance of HC N plotted as afunction of the dust temperature in each pixel of G5.899-0.429. X ( HC N ) x - T dust (K) Figure 8.
Top: The HCO + (1-0) spectra superimposed on the calculated HC N abundance map of G12.804-0.199. Bottom: Abundance of HC N plotted as afunction of the dust temperature in each pixel of G12.804-0.199. dust (K) X ( HC N ) x - Figure 9.
Top: The HCO + (1-0) spectra superimposed on the calculated HC N abundance map of G326.641+0.612. Bottom: Abundance of HC N plotted asa function of the dust temperature in each pixel of G326.641+0.612. dust (K) X ( HC N ) x - Figure 10.
Top: The HCO + (1-0) spectra superimposed on the calculated HC N abundance map of G327.293-0.579. Bottom: Abundance of HC N plotted asa function of the dust temperature in each pixel of G327.293-0.579. dust (K) X ( HC N ) x - Figure 11.
Top: The HCO + (1-0) spectra superimposed on the calculated HC N abundance map of G337.916-0.477. Bottom: Abundance of HC N plotted asa function of the dust temperature in each pixel of G337.916-0.477. igure 12. Top: The HCO + (1-0) spectra superimposed on the calculated HC N abundance map of G345.448+0.314. Bottom: Abundance of HC N plotted asa function of the dust temperature in each pixel of G345.448+0.314. (cid:0)✁✂✄☎✆✝✞✆✄✁✟✟ RCW 97
Figure 13.
Top panels: The averaged spectra of HC N (10-9) in G12.804-0.199 and RCW 97 where its abundance is highest. Gaussian fits to individualcomponents are plotted using solid red lines and the cumulative fits with solid green lines. Vertical dashed lines mark the systemic velocity of each sourcemeasured from the Gaussian fit to their N H + (1-0) lines. Bottom panels: The abundance of HC N plotted as a function of its nonthermal line width inG12.804-0.199 and RCW 97. ✁✂✄☎☎✆✝✂✞✟☎ (cid:0)✠✟✂✄✝✞✆✝✂✠☎☎(cid:0)✡✞✁✂✞✞✄☛✝✂✡✠✞☞✡✌✍✎✏ ☎✁ ✍✎ ✏ ☎✑ Figure 14.
The relationship between N ( N H + )/ N ( HC N ) and T dust in all of our sources. For RCW 95 and S36, the column densities of N H + are fromYXW18. For other sources, we calculated their column densities of N H + in the same way.in the same way.