THE SECOND SURVEY OF THE MOLECULAR CLOUDS IN THE LARGE MAGELLANIC CLOUD BY NANTEN. II. STAR FORMATION

A survey of the molecular clouds was carried out in ¹²CO (J = 1–0) by NANTEN, a 4 m radio telescope of Nagoya University at Las Campanas Observatory, Chile (Paper I). The observed region is about 30 deg² and covers the region where the CO emission was detected by the NANTEN first survey (e.g., Fukui et al. 1999; Mizuno et al. 2001). The observed grid spacing was 2', corresponding to ∼30 pc at a distance of the LMC, 50 kpc, with a 26 half-power beam width at 115 GHz. The spectral intensities were calibrated by employing the standard room-temperature chopper wheel technique (Kutner & Ulich 1981). An absolute intensity calibration was made by observing Orion-KL (R.A. (B1950) = 5^h32^m470, decl. (B1950) = −5°24'21'') by assuming its absolute temperature, T*_R, to be 65 K. The rms noise fluctuations were about 0.07 K at a velocity resolution of 0.65 km s⁻¹ with about 3 minutes' integration for an on-position. The typical 3σ noise level of the velocity-integrated intensity was about 1.2 K km s⁻¹ (Figure 1).

Zoom In Zoom Out Reset image size

Fukui et al. (2008) identified 272 molecular clouds of which 230 are detected at more than two observing positions (hereafter "GMCs," in this paper) by using cloud identifying algorithm, cprops (Rosolowsky & Leroy 2006). The radius and virial mass of the clouds range from 10 to 220 pc, and 9 × 10³ to 9 × 10⁶M_☉, respectively. The CO luminosity and virial mass of the GMCs show a good correlation, and a conversion factor, X_CO, from a CO intensity to an H₂ column density was derived to be (7 ±2) ×10²⁰ cm⁻² (K km s⁻¹)⁻¹ by assuming virial equilibrium. The sensitivity in N(H₂), then, corresponds to N(H₂) = 8 × 10²⁰ cm⁻² and the range of mass of the GMCs is 2 × 10⁴M_☉ to 7 × 10⁶M_☉, respectively. The details of the observations and the method to identify the GMCs are found in Paper I and Rosolowsky & Leroy (2006).

In order to identify the current massive star and cluster forming molecular clouds, H ii regions and young stellar clusters were searched for in the published catalogs. The association of these objects and the molecular clouds were determined.

Henize (1956) and Davies et al. (1976) cataloged more than 300 Hα emission nebulae in the LMC. The estimated diameters of the H ii regions range from ∼10 pc to ∼400 pc. In addition, an extensive Hα photometry of the 240 H ii regions was carried out by Kennicutt & Hodge (1986). Figure 2 shows luminosity distribution of these H ii regions, indicating that the current sample of H ii regions includes those with Hα luminosities as faint as ∼10³⁶ erg s⁻¹ at the faint end. This shows that the sensitivity of the survey is high enough to detect the Orion nebula, ∼4 × 10³⁶ erg s⁻¹, (Gebel 1968) at the distance of the LMC. The Orion Nebula is ionized primarily by a single O7 star together with at least B0.5, B3, and B0.5 (O9.5) stars (Goudis 1982). This faint end of the luminosities is comparable to those of small Galactic H ii regions that can be ionized by a single O9 star (e.g., Panagia & Ranieri 1973). Detailed studies of several faint H ii regions in the LMC show that most of the H ii regions are ionized primarily by a star as massive as B0 accompanied by several other massive but cooler stars, except for two ionized by a single mid-O star and B0 star (Wilcots 1994). It is also suggested that the faint H ii regions in the catalogs are ionized by a single O star and those luminous ones (L ≳ 5 × 10³⁷ erg s⁻¹) are by OB associations (Kennicutt & Hodge 1986).

Zoom In Zoom Out Reset image size

A mosaic image of the 1.4 GHz continuum emission observed with the Parkes Telescope (Filipovic et al. 1995) and the ATCA are combined to present the thermal and nonthermal radio emission of the LMC covering 108 × 123 (e.g., Filipovic & Staveley-Smith 1998; Hughes et al. 2007). This image was also used to determine if there are any H ii regions escaped from optical identification due to the extinction by molecular clouds. The angular resolution of the data is 40'', corresponding to ∼10 pc at a distance of the LMC. Thus one should keep in mind that a size of the H ii regions of this study is larger than ∼10 pc.

The LMC has more than 6500 stellar clusters and OB associations (Bica et al. 1999). Photometric properties of about 10% of the brightest ones were measured to derive their masses and ages (Bica et al. 1996). In this paper, we use these clusters with known properties to study the formation of cluster and OB association in the molecular clouds. These 120 OB associations and 504 stellar clusters are classified into 10 types of different ages (SWB0–VII; Bica et al. 1996) from their U − B and B − V indices. In Table 1, the numbers of these objects according to their ages in the area covered with the second NANTEN survey are shown. We shall call these objects "clusters" throughout this paper and we should note that this includes both clusters and OB associations.

Figures 1, 3(a), and (b) show the distribution of the Hα emission (Kim et al. 1999), young clusters (younger than 10 Myr), and clusters older than 10 Myr together with the molecular clouds, respectively.

Zoom In Zoom Out Reset image size

Figure 1 shows that the overall distribution of the Hα local peaks and GMCs are well coincident to each other, while the extent of the individual Hα emitting regions and the size of the GMCs are different. Almost all the luminous H ii regions with high flux densities, like 30 Dor to N 159, N 11, N 63/N 64 complexes, N 44 and N 206 are associated with GMCs, while fainter and diffuse extended Hα emission show less degrees of association. Some luminous H ii regions, like N 51, are associated with only small molecular clouds, which may indicate molecular gas dissipation after formation of massive H ii regions. It is also noted that there are several GMCs not associated with Hα emission (see also Section 4.1)

Figure 3(a) shows that young clusters are often found at, or near the peak of the GMCs. Some of the young clusters associated with GMCs are forming groups with a few to ten clusters. These active cluster forming regions, such as N 159, N 11, and N 44, are found especially toward bright H ii regions. Note that a number of clusters in the northeastern region in the LMC and some in the west of 30 Dor are isolated from molecular clouds. These groups of clusters are located inside large cavities of Hα emission (see also Figure 1), two of the supergiant shells, LMC 4 and LMC 3, respectively (Meaburn 1980).

Compared with the youngest group of clusters, the older clusters have less degree of association with the GMCs (Figure 3(b)). The second youngest group of clusters, SWB I, is presented in blue crosses in Figure 3(b), showing that the correlation with the molecular clouds is low. Nevertheless there are several regions where the number of SWB I clusters is enhanced, for example, inside a supergiant shell, LMC 4, and the south of 30 Dor. It is interesting to note that the SWB 0 clusters are also gathered in these regions, suggesting that the distribution of SWB 0 and I clusters still retain the information of their formation sites. On the other hand, the older clusters, SWB II–VII, are distributed more uniformly over the galaxy showing no trend of association with the GMCs. It is to be noted here that the older clusters are distributed more widely over the galaxy compared with the region where the current star formation is observed (Bica et al. 1996)

The overall distribution of H ii regions, young clusters, and older clusters indicates that the GMCs are the sites of current massive star formation and of cluster formation. To show the associated object more clearly, we present close-up views of the GMCs and the young objects in Figures 4 and 5(a)–(f) (Figure 4 as a guide). These panels confirm that the H ii regions and young clusters are often associated with molecular clouds well, while the older clusters are distributed randomly.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

In the following, we report some individual active star-forming regions in detail. In the N 11 (DEM 41) region, the young clusters are found in a bright, large complex of H ii regions, while they are not exactly found toward the GMCs (Figure 5(a)). Only small clouds are found at the outer edge of N 11 nebulae with young clusters near the edge of the molecular clouds. The cluster at the center of N 11, LH 9, is neither associated with Hα emission peaks nor molecular clouds. This suggests that the parent cloud of LH 9 has been dissipated, and that triggered star formation occurred at the outer region as discussed by, for example, Israel et al. (2003a). A group of H ii regions are found near the western edge of the bar in the N 79/N 83 region (Figure 5(b)). In contrast with the N 11 region, young clusters associated with the GMCs and bright Hα emission are found near the peak of the GMCs in this region, and the young clusters without a bright H ii region are at the center of diffuse shell-like Hα emission. In Figure 5(b), there are a few GMCs hosting no H ii regions, one of which is at the western end of the N79/N 83 region.

Small groups of young clusters are found near the peak of a massive GMC in N 44 (DEM158, 160, 166, 167, 169) as well as small GMCs along the supergiant shell LMC 4 (Figure 5(c)). On the other hand, only one or two clusters are associated with a GMC near the LMC bar (Figure 5(d)). One may speculate that clusters are formed in groups in the massive GMCs of M ∼ 10⁶M_☉ or near a supergiant shell. It is also noted that a bright H ii region, the N 51 complex, is associated with a very small molecular cloud, maybe a remnant of the parent cloud of this bright H ii region.

Figure 5(f) shows the 30 Dor (DEM 263) region to the molecular ridge including active star-forming site, N 159 (DEM 271, 272), and the Arc region. The most remarkable feature in Hα emission is a bright complex of H ii regions in 30 Dor, while massive molecular clouds are found not exactly toward 30 Dor but extending to the south as already noted by several authors (e.g. Cohen et al. 1988; Indebetouw et al. 2008; Paper I). Most of the youngest clusters of τ< 10 Myr are associated with the GMCs or found near the GMCs. On the other hand, the clusters of 10 <τ< 30 Myr are away from the GMCs and mostly found between 30 Dor and N 159 where the bright Hα emission from 30 Dor is extended but only small molecular clouds are detected. The most active current star formation site in this figure is a well known region, N 159, where young clusters as well as a number of H ii regions are found at or near the peaks of the molecular ridge. The detailed studies of star formation activities and molecular clouds are carried out by several authors (e.g., Johansson et al. 1998; Minamidani et al. 2008; Indebetouw et al. 2008; Mizuno et al. 2009). In the south of N159, only smaller H ii regions are associated with the molecular ridge. These results indicate that the southern region may be younger than the north.

There are several other regions with active star formation in Figure 5(f), N 148, N 180, N 206, and N 214, where groups of H ii regions, some of which contain a young cluster, are associated with GMCs. It is interesting to note that these groups of H ii regions are mostly found off from the peak of the GMCs. It should be noted that the NANTEN beam is not capable of resolving the individual local peaks of the GMC and the parent core of the group of H ii regions may have escaped from detection. Nevertheless, the positional offset of the peak of the GMC and the group of H ii regions indicate the dissipation of the molecular gas by active star formation. It is also to be noted that there are several GMCs, especially at southern edge of this figure without significant Hα emission or clusters.

Figure 6 shows the distribution of the projected separations of H ii regions and stellar clusters from the nearest CO emission with an integrated intensity above 1.2 K km s⁻¹ (3σ noise level). The lines in Figure 6 represent the frequency distribution expected if the same number of the H ii regions or clusters are distributed at random in the observed area. It is clearly shown by eye that the distribution of the youngest clusters with an age smaller than 10 Myr, i.e., SWB 0 (Bica et al. 1996), and the H ii regions are sharply peaked within 100 pc of CO emission exhibiting strong spatial correlations. On the other hand, the older clusters, SWB I (τ>10 Myr) or older, show much weaker or no correlation. This figure again indicates that the H ii regions and SWB 0 clusters are well associated with molecular clouds as well as that rapid cloud dissipation after cluster formation.

Zoom In Zoom Out Reset image size

Hereafter, we will focus on the H ii regions and SWB 0 clusters in the discussion and call SWB 0 cluster as "young cluster" unless otherwise stated.

3.2.1. Determination of association of the H ii regions and young clusters

3.2.2. Determination of Association of Radio Continuum Sources toward the Molecular Clouds without Optically Identified H ii regions

It was shown in Fukui et al. (1999) that the GMCs can be classified into three groups based on a sample of 55 GMCs with a mass ranging from 2 × 10⁵M_☉ to 3 × 10⁶M_☉: (1) starless GMCs; (2) those with small H ii regions whose Hα luminosity is less than 10³⁷ erg s⁻¹; and (3) those with stellar clusters and large H ii regions of Hα luminosity greater than 10³⁷ erg s⁻¹.

The current comparison of the molecular clouds with the clusters and H ii regions gives a consistent result. Here we classify the molecular clouds into three types according to the association with massive star formation activities:

• 1.  
  starless molecular clouds in the sense that they are not associated with H ii regions or young clusters (Type I);
• 2.  
  molecular clouds with H ii regions (Type II); and
• 3.  
  molecular clouds with H ii regions and young clusters (Type III).

It should be noted that "starless" here means without star-forming activities with stars more massive than early O star capable of ionizing H ii regions, and it does not exclude the possibility of associated young low-mass stars. Comparisons of the GMCs with young, low or intermediate mass stars are now possible by using recent results by the IR satellites, like Spitzer (e.g., Mexiner et al. 2006, "Surveying the Agents of a Galaxy's Evolution") and AKARI (e.g., Ita et al. 2008; Murakami et al. 2007). Comparisons of these YSOs (Whitney et al. 2008) and the GMCs are found elsewhere (Indebetouw et al. 2008; T. Onishi et al. 2009, in preparation).

Table 2 lists the associated H ii regions and young clusters for each molecular cloud. Out of 272 molecular clouds, 72, 142, and 58 are found to be Type I, II, and III, respectively (see also Table 3). Figures 7–9 present the examples of the molecular clouds of each type. Examples are chosen from the most massive clouds from each type. It is interesting to note that the most massive Type I molecular clouds have similar size and mass to those of Type II, while the number of massive Type I is less. To study the physical properties of the molecular clouds, one has to keep in mind that the completeness limit of the NANTEN survey of M_CO = 5 × 10⁴M_☉. Table 3 also summarizes a number of GMCs with M_CO > 5 × 10⁴M_☉ for each type. Out of 191 GMCs, 46, 96, and 49 GMCs are found to be Type I, II, III, showing that about a half of them are Type II, and a quarter is Type I and Type III, respectively.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Figures 10(a)–(f) show the radial distribution of CO emission; the number and the surface density, Σ, of the molecular clouds with Types I, II, and III, respectively. The surface density, Σ, is derived by integrating the CO luminosity within annuli spaced by 4' and then divided by an area of the annuli. The center used is α(J2000) = 5^h176, δ(J2000) = −69°2' determined from the kinematics of the H i observations by Kim et al. (1998). To see the angular distribution of the CO emission, the distribution of the clouds and the surface density, Σ, for each molecular cloud type are also presented in Figures 11(a)–(f). Here, the surface density, Σ, is derived by integrating CO luminosity over a sector with a 20° width and then divided by an observed area of the sector. The CO luminosity to mass conversion is carried out by assuming a conversion factor, X_CO of 7 × 10²⁰ cm⁻² (K km s⁻¹)⁻¹ (Paper I) for both Figures 10 and 11.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Figure 10 shows that the radial profile of the surface density decreases moderately along the galactocentric distance for Type II as is also seen in the nearby spiral galaxies (e.g., Wong & Blitz 2002), while those for Types II and III are rather flat with respect to the radial distance. It is interesting to note that the number distribution and surface density show different radial profiles for Type I; the number increases at large radial distances but the surface density is relatively constant. This indicates that the more massive Type I GMCs are found at the large radial distances. It is also notable that there is a sharp enhancement of the number of the clouds around 1.5 kpc for Types II and III. This enhancement is due to the molecular ridge, N11, and N44. This enhancement is also seen in the angular distribution, especially at about 120° due to the molecular ridge.

Cluster forming clouds provide us with precious information on understanding physical processes of cluster formation. We shall examine the physical properties of the GMC types. In this section, only the GMCs with mass M_CO > 5 × 10⁴M_☉ are considered. Figure 12 shows the distribution of the line width, size, and mass of the GMCs, respectively. The upper panels are the GMCs not associated with the H ii regions and the clusters, i.e., those with no sign of massive star formation (Type I). The middle ones are the GMCs associated only with the H ii regions, i.e., those showing massive star formation but no cluster formation (Type II). The lower are the GMCs associated with the H ii regions and clusters, i.e., those actively forming stars and clusters (Type III). Table 4 summarizes a mean and standard deviation of the line widths, sizes, and masses of each GMC type.

Zoom In Zoom Out Reset image size

Table 4. Physical properties of the GMCs

		Line Width		Size		Mass
GMC Type	Number of Clouds	〈ΔV〉 (km s−1)	σΔV (km s−1)	〈R〉 (pc)	σR (pc)	〈MCO〉 (105 M☉)	$\sigma _{M_{\rm CO}}$ (105 M☉)
I	46(24%)	5.9	2.5	40	15	2	2
II	96(50%)	5.5	2.2	36	19	3	4
III	49(26%)	7.0	3.0	46	30	6	20

Notes. Properties of the GMCs with mass, M_CO > 5 × 10⁴M_☉. The average values of line width, size, and mass for each type are shown in 〈 〉. The average and σ of the size are derived for 164 GMCs with minor axis more than the NANTEN beam (see also Paper I).

Download table as:  ASCIITypeset image

Figure 12 and Table 4 indicate that a significant difference is not seen in the line width and size among the three types, while the mass distribution shows a slight difference. The frequency distribution of mass of Types I and II shows a peak at M_CO ∼ 10⁵M_☉ and then decreases as the mass becomes large, while that of Type III is rather flat and the average mass of Type III is larger than that of Types I and II. The small difference in the physical properties among three types suggests that the integrated properties of the GMCs at a scale of 30–100 pc are not so sensitive to the local star formation activities until the last stage of cloud dissipation.

Almost all the GMCs in the solar vicinity are forming massive stars actively as indicated by associated H ii regions and/or OB associations in addition to a number of young low-mass stars (Dame et al. 1987; Blitz 1993 and references therein). There are only two GMCs which show no signs of massive star formation in the solar neighborhood among ∼20 GMCs; the Maddalena's cloud (Maddalena & Thaddeus 1985) and ON-1 cloud complex (Israel & Wootten 1983). The reason for not forming massive stars may be either it is at a very young stage prior to massive star formation (Maddalena & Thaddeus 1985; Israel & Wootten 1983) or that it is in a late stage after active star formation (Lee et al. 1994).

A large number of starless GMCs in the LMC suggests that the timescale in star formation is significantly longer in the LMC than in the Galaxy. The ionization degree in a molecular cloud is likely determined by the far-ultraviolet (FUV) photons of stellar radiation fields (McKee 1989; Nozawa et al. 1991). In the LMC, the FUV flux is several times higher (Israel et al. 1986) and the dust extinction is smaller by a factor of 3–4 for a given gaseous column density (Koornneef 1982). Since the timescale of the diffusion of magnetic field is proportional to the ionization degree (Spitzer 1978), the contraction of cloud may be slowed down by the magnetic field. In addition, the cooling rate via molecular and dust emission is expected to be smaller in the LMC than in the Galaxy, helping star formation activity to slow down. Higher ionization degree and smaller cooling rate are basically the consequences of lower metallicity in the LMC (Dufour 1984) and are both likely to support for the retarded star formation.

An alternative idea to explain starless GMCs is that the GMCs in the LMC are of very recent formation, a situation possibly similar to the Maddalena's clouds and ON-1 cloud complex. It is well known that the LMC has a number of supershells expanding to accumulate the interstellar matter (Meaburn 1980; Oey 1996; Kim et al. 1999). Yamaguchi et al. (2001a) investigated the possible correlation between GMCs and supergiant shells and concluded that one third of GMCs may be located toward the shell boundaries, suggesting that a significant number of GMCs may have been formed under the triggering by expanding shells. The spatial distribution of the starless GMCs is however fairly random, showing little correlation with supergiant shells. It seems therefore the retarded star formation in the LMC is not due to some local environment or dynamical activities.

Finally, we shall comment on a possible link between Type I GMCs and the formation of populous stellar clusters. It is tempting to speculate that there is a link between populous clusters and the Type I GMCs. A possible explanation is that a longer timescale of star formation allows the formation of protocluster molecular condensations as massive as 10⁵M_☉, which can lead to form populous clusters. This will never happen in the Galaxy because of the star formation immediately after formation of protocluster condensations having mass of 10³M_☉.

The molecular clouds are considered to be formed in neutral gas, and as they evolve, they form stars and clusters, being dissipated by stellar winds or UV radiation from the massive stars or supernova explosions. At the end of their life, the newly formed stars and clusters remain. The process is, however, still unclear quantitatively, because we need a complete data set of molecular clouds to estimate the evolutionary timescale statistically. In our Galaxy, it is difficult to obtain such a complete sample due to the heavy contamination toward the Galactic plane. On the other hand, a face-on galaxy like the LMC is suitable for collecting a complete sample, which enables us to investigate the evolutionary process more quantitatively. By using the complete data set of the GMCs for a whole galaxy, we shall estimate their evolutionary timescale. In this section, only the GMCs with mass M_CO > 5 × 10⁴M_☉ are considered.

The evolutionary sequence of the GMCs is schematically drawn in Figure 13 together with examples of the GMCs corresponding to each stage. In the first stage, the GMC Type I, the GMCs show no sign of massive star formation. The second stage, the GMC Type II, is the GMCs associated only with H ii regions. They are forming massive stars, but clusters have not appeared yet. The third, the GMC Type III, is the GMCs associated with both H ii regions and clusters. They are actively forming clusters. Molecular gas around newly formed clusters is partly dissipated, but the GMCs are still massive as is seen in Figure 12. The last is when the GMCs have been completely dissipated and only the young clusters and/or SNRs are found.

Zoom In Zoom Out Reset image size

If we assume that the GMCs and clusters in the LMC are being formed nearly steadily, we can estimate each evolutionary timescale according to the above classification. First, we shall estimate the timescale of the cloud dissipation. We found that ∼66% of the youngest clusters of τ< 10 Myr are associated with the GMCs (see Section 3.2). If we assume that the clusters in the LMC are being formed nearly steadily in the past 10 Myr, this result means the GMCs can survive during ∼66% of the cluster age, 10 Myr, and are dissipated in a few Myr after formation of clusters due to the UV photons from the clusters. The timescale for the GMC Type III is thus considered to be ∼7 Myr. If we further assume that the massive star formation and cluster formation occur nearly steadily, the timescale for each stage is proportional to the number of the GMCs. Accordingly, the timescales for the GMC Type I and II are estimated to be 6 Myr and 13 Myr, respectively. As a result, the typical lifetime of the GMCs, corresponding to the total lifetime of the GMC from Type I to III, is roughly ∼30 Myr.

Fukui et al. (1999) and Yamaguchi et al. (2001b) made a comparison of 55 GMCs with mass above 2 × 10⁵M_☉ from the NANTEN first survey with the young objects. Their result shows that 6 GMCs show no sign of massive star formation, 9 are associated with H ii regions, and 28 with young clusters, indicating that the population of Type III GMCs is the highest, 50%, and Type II and I are 21% and 14%, respectively. The higher sensitivity of the second survey made us possible to obtain a sample of the GMCs with mass as small as 5 × 10⁴M_☉. This provides us with a complete sample of the GMCs, which are massive enough to produce massive stars. The numbers of Type I and Type II GMCs are increased in the current work compare to the first survey more than that of the Type III, because the mass of the Type I and Type II are smaller than the mass of the Type III, and thus, the higher sensitivity of the survey increased the number of Type I and Type II more. As a consequence of it, the estimated timescale of the a GMC now becomes longer from the previous result.

We summarize the evolutionary timescale of the GMCs in Table 3. The GMCs form massive stars, and H ii regions appear after ∼6 Myr from their birth. After 13 Myr, clusters start to be formed, then also start dissipating the surrounding molecular gas. The GMCs continue to form clusters actively, being dissipated by the UV photons and stellar winds from the clusters. After ∼7 Myr, the GMCs have been almost dissipated by the newly formed clusters, and eventually, by supernova explosions. The above time estimation should include an error. The age determination for the clusters contains uncertainty, but this changes only the absolute timescale for each stage. Other errors are possible due to a simple assumption of the constant formation rate for the clusters and the GMCs. Our estimation, nevertheless, shows a typical evolutionary sequence of the GMCs. Further quantitative detailed studies, such as a comparison with Hα flux, determining the age and mass of the clusters, and observations at higher resolution, will lead us to better understandings of star formation processes and the cloud dissipation in the LMC. Furthermore, detailed comparisons of the molecular clouds with H i give us a clue to understand the molecular clouds formation. These studies are found in elsewhere (e.g., Wong et al. 2009; Fukui et al. 2009)