New Look at the Molecular Superbubble Candidate in the Galactic Center

The $l\!=\!+1.\!\!^\circ3$ region in the Galactic center is characterized by multiple shell-like structures and their extremely broad velocity widths. We revisit the molecular superbubble hypothesis for this region, based on high resolution maps of CO {\it J}=1--0, $^{13}$CO {\it J}=1--0, H$^{13}$CN {\it J}=1--0, H$^{13}$CO$^{+}$ {\it J}=1--0, SiO {\it J}=2--1, and CS {\it J}=2--1 lines obtained from the Nobeyama radio observatory 45-m telescope, as well as CO {\it J}=3--2 maps obtained from the James Clerk Maxwell telescope. We identified eleven expanding shells with total kinetic energy and typical expansion time $E_{\rm kin}\!\sim\! 10^{51.9}$ erg and $t_{\rm exp}\!\sim\! 10^{4.9}$ yr, respectively. In addition, the $l\!=\!+1.\!\!^\circ3$ region exhibited high SiO {\it J}=2--1/H$^{13}$CN {\it J}=1--0 and SiO {\it J}=2--1/H$^{13}$CO$^{+}$ {\it J}=1--0 intensity ratios, indicating that the region has experienced dissociative shocks in the past. These new findings confirm the molecular superbubble hypothesis for the $l\!=\!+1.\!\!^\circ3$ region. The nature of the embedded star cluster, which may have supplied 20--70 supernova explosions within 10$^5$ yr, is discussed. This work also show the importance of compact broad-velocity-width features in searching for localized energy sources hidden behind severe interstellar extinction and stellar contamination.


INTRODUCTION
The Galactic center contains a large quantity of warm (T k = 30-60 K) and dense [n (H 2 ) ≥ 10 4 cm −3 ] molecular gas (Morris et al. 1983;Paglione et al. 1998). Gases in the central molecular zone (CMZ; Morris & Serabyn 1996) exhibit highly turbulent and complex kinematics with large velocity dispersions (Oka et al. 1998b). In addition to the ubiquity of shock-origin molecules (Hüttemeister et al. 1998;Requena-Torres et al. 2006), the highly turbulent kinematics of molecular gas in the CMZ can be attributed to the release of kinetic energy by numerous supernova (SN) explosions. It has been suggested that the instantaneous and repetitive phases of star formation have dominated continuous star formation in the CMZ in the past. Despite the abundance of dense molecular gas, star formation is currently inactive in the CMZ. The current star forma-tion rate (SFR) ∼ 0.04-0.1 M ⊙ yr −1 (Yusef-Zadeh et al. 2009;Immer et al. 2012) is at least an order of magnitude lower than that expected from the amount of dense gas in the CMZ (Kruijssen et al. 2014).
In contrast, two well-known young super star clusters in the CMZ, the Arches and Quintuplet clusters, are considered to be formed by microbursts of star formation several Myr ago (Figer et al. 1999a(Figer et al. , 2002. Recent gamma-ray observations have revealed a pair of huge shell-like structures called "Fermi bubbles," extending up to 50 • above and below the Galactic center. The "Fermi bubbles" are considered to have been formed by a pair of energetic jets from Sgr A* or a nuclear starburst in the last several tens of Myr (Su et al. 2010). In addition, dozens of high-velocity compact clouds (HVCCs) with large kinetic energy were detected in our recent studies on the CMZ (Oka et al. 1998b(Oka et al. , 1999(Oka et al. , 2001(Oka et al. , 2012; Tanaka et al. 2014;Takekawa et al. 2017Takekawa et al. , 2019b. Some of the HVCCs exhibit one or more shell-like structures that may have been accelerated by a series of SN explosions that occurred in massive star clusters 10-30 Myr ago Tsujimoto et al. 2018). The presence of these shells suggests that numerous micro-starbursts have occurred in the CMZ 10-30 Myr ago. The intense 6.7 keV ionized iron emission line widely detected in the CMZ via Galactic center diffuse X-ray (GCDX) was reported to be produced in hot plasma, whose possible energy source is 10 2-3 SN explosions at a rate of 10 −2 -10 −3 yr −1 (Yamauchi et al. 1990; Koyama et al. 2007c). This GCDX also supports the past active star formation scenario.
In this paper, we present newly obtained molecular line datasets of the l = +1. • 3 region as parts of a largescale survey of the CMZ ( §2). The new CO maps show the detailed spatial and velocity distribution of this region, revealing multiple expanding shells. In addition, the distribution and kinematics of shocked gas are discussed based on the SiO line data ( §3). Based on these datasets, we revisit the molecular superbubble scenario for the l = +1. • 3 region ( §4). Then we summarize this work at the last section ( §5). The distance to the Galactic center is assumed as D = 8.3 kpc in this paper.

CO J=1-0 Line
We observed the 12 CO J=1-0 (115.27120 GHz) line on 19-29 January 2011, and the 13 CO J=1-0 (110.20135 GHz) line on 26-31 January, 1-15 February, and 9-23 March 2016 using the NRO 45-m telescope . The target areas were set to −0. • 8 ≤ l ≤ +1. • 4 and −0. • 35 ≤ b ≤ +0. • 35 for the 12 CO observations, and to −1. • 4 ≤ l ≤ +1. • 4 and −0. • 35 ≤ b ≤ +0. • 35 for the 13 CO observations. The 25-beam array receiver system (BEARS; Sunada et al. 2000;Yamaguchi et al. 2000) and AC45 spectrometer (Sorai et al. 2000) system were employed in the 12 CO line observations. Moreover, the four-beam receiver system on the 45-m telescope (FOREST; Minamidani et al. 2016) with the spectral analysis machine on the 45m telescope (SAM45; Kuno et al. 2011;Kamazaki et al. 2012) system were used in the 13 CO line observations. The half-power beamwidths (HPBW) of the telescope with BEARS and FOREST were approximately 15 ′′ at 115 and 110 GHz, respectively. We used the AC45 spectrometer in the 500 MHz bandwidth (0.5 MHz resolution) mode and SAM45 spectrometer in the 1 GHz (244.14 kHz resolution) mode. The typical noise temperature (T sys ) was ∼ 800 K during the 12 CO line observations, and ranged from 150 K to 300 K during the 13 CO line observations. All NRO 45-m datasets were reduced on the NOSTAR 1 reduction package. We used linear, or if necessary, the lowest degree polynomial fittings to subtract the baselines from the obtained spectra. Less than 1% of spectra needed non-linear baseline subtraction. The data were spatially convolved using Bessel-Gaussian functions and resampled onto an 7. ′′ 5 × 7. ′′ 5 × 2 km s −1 grid. The temperature scale of the 12 CO J=1-0 line data were determined by comparing these data with the previous 12 CO J=1-0 line data (Oka et al. 1998b), whose intensity scale was calibrated to the radiation temperature scale of the Harvard-Smithsonian Center for Astrophysics survey (Dame et al. 1987). The 13 CO J=1-0 line data in antenna temperature (T * a ) scale were converted to the main-beam temperature (T MB ) scale by multiplying by the main-beam efficiency (η MB ), whose value was measured to be 0.43 at 110 GHz.

NRO 45-m Large Program
The H 13 CN J=1-0 (86.33986 GHz), H 13 CO + J=1-0 (86.75429 GHz), SiO J=2-1 (86.84696 GHz), and CS J=2-1 (97.98096 GHz) line observations were performed in the NRO 45-m Telescope Large Program: "Complete Imaging of the Dense and Shocked Molecular Gas in the Galactic Central Molecular Zone;" it was approved for 2018-2019 and 2019-2020 (Takekawa et al. in preparation). The details of the observations and data reduction are summarized in the forthcoming paper. We mapped the area of −1. • 5 ≤ l ≤ +1. • 5 and −0. • 25 ≤ b ≤ +0. • 25 with the FOREST + SAM45 sys-tem in January-May 2019 and January-April 2020. The HPBW of the telescope was ≃ 19 ′′ at 86 GHz. The SAM45 spectrometer was employed in the 1 GHz bandwidth (244.14 kHz resolution) mode. During the H 13 CN, H 13 CO, SiO, and CS line observations, T sys ranged from 150-300 K. The obtained datasets were reduced on the NOSTAR reduction package. We subtracted the baselines from the spectra by fitting the linear lines. The data were spatially convolved using Bessel-Gaussian functions and resampled onto an 7. ′′ 5 × 7. ′′ 5 × 1 km s −1 grid. We converted the T * a scale into the T MB scale with η MB = 0.49 at 86 GHz.

12 CO J=3-2 Line
The JCMT Galactic plane survey team performed the 12 CO J=3-2 line (345.795990 GHz) observations of the CMZ using JCMT in July-September 2013, July 2014, and March-June 2015 (14 h in total; Parsons et al. 2017;Eden et al. 2020). These data were obtained with the heterodyne array receiver program (HARP; Buckle et al. 2009) and auto correlation spectral imaging system (ACSIS). The HPBW of the telescope was approximately 14 ′′ at 345 GHz. The ACSIS spectrometer was operated in the 1 GHz bandwidth (976.56 kHz) mode. During the HARP observations, T sys typically ranged from 100-200 K. In this paper, we use the 12 CO J=3-2 data for the entire l = +1. • 3 region. We reduced the JCMT data with the Starlink 2 software package. The data were smoothed with Gaussian functions and resampled on a 7. ′′ 5 × 7. ′′ 5 × 2 km s −1 grid to obtain the final map. We converted the T * a scale into the T MB scale with η MB = 0.64 at 345 GHz. Figure 2 shows the l-b maps of the 12 CO and 13 CO J=1-0; SiO and CS J=2-1; and 12 CO J=3-2 line emissions integrated over a velocity range of 80 km s −1 ≤ V LSR ≤ 180 km s −1 . The 12 CO, SiO, and CS line emissions exhibit similar morphologies, containing two ellipses with central emission cavities centered at (l, b) ∼ (+1. • 27, 0. • 00), (+1. • 24, +0. • 10). Hereafter, we call these the southern and northern shells, respectively. The position of the northern shell coincides approximately with that of the "minor shell" described in Oka et al. (2001) and the C/C 1 shells described in Tanaka et al. (2007). The position of the southern shell coincides with the major shell/shell A described in Oka et al. (2001) and Tanaka et al. (2007). An "arc" structure, hereafter referred to as the Arc, was detected in the 12 CO, SiO, and CS maps, which is elongated from the bottom part of the southern shell at (l, b) ∼ (+1. • 24, −0. • 04), reaching (l, b) ∼ (+1. • 22, −0. • 14). In the 13 CO map, the northern shell is fainter, and the Arc is brighter and more spatially extended than those in the 12 CO, SiO, and CS maps. The Arc traces the northwestern edge of shell B described in Tanaka et al. (2007). Figure 3 shows the composite velocity channel maps of the l = +1. • 3 region. At lower velocities (V LSR ≤ 100 km s −1 ), the 12 CO J=1-0 line emission dominates the color of the channel maps. At higher velocities (V LSR ≥ 100 km s −1 ), where the northern and southern shells appear clearly, the 12 CO J=3-2 line emission is more intense than the 12 CO J=1-0 line. The clear ellipses (V LSR = 117-153 km s −1 ) and deep emission holes (V LSR = 75-99 km s −1 ) of the southern shell are also ev-2 http://starlink.eao.hawaii.edu/starlink ident. In the northern shell, the distinct arcs can also be observed at V LSR = 129-159 km s −1 .

Spatial Distribution
Inspecting the CO J=3-2 and SiO J=2-1 datasets comprehensively, we identified 11 shell-like structures. In general, their spatial sizes change gradually with increasing velocity to form topologically closed threedimensional structures. We named these 11 shells as follows: those around the northern shell as N1-7, those around the southern shell as S1-3, and the one adjacent to the Arc as B following Tanaka et al. (2007). The SiO J=2-1 line emission is prominent at the edges of these shells. The spatial and velocity behavior of the northern and southern shells are represented by N5 and S2, respectively. The outlines of the identified shells are shown in the top right panel of Figure 2. Among the 11 shells, 9 correspond to those previously identified in Oka et al. (2001) and Tanaka et al. (2007). The center positions (l, b), sizes (∆l, ∆b), and names of the corresponding shells described in previous works are listed in Table 1.

Spatial-Velocity Distribution
We present the b-V maps of the 12 CO J=3-2 and SiO J=2-1 lines in Figure 4. The latitude-velocity (b-V) behavior of CO and SiO emissions are similar in the l = +1. • 3 region, while the SiO emission favors edges and both high-velocity ends of the shells. The SiO emission is also enhanced at overlapping areas between the shells. The N1-7 and S1-2 shells appear as elliptical or arclike structures with central cavities in these b-V maps. Such an ellipsoidal shape in the l-b-V space strongly suggests the kinematics of an "expanding" shell. The systemic velocity (V sys ) and expansion velocities (V exp ) of the shells were determined by eye, except for N3, N5, and S2 (see Section 4.3). These are listed in Table 1.

Classification of Identified Shells
Highly complex molecular gas distribution and kinematics in the l = +1. • 3 region prevents probing the shells deeply. Thus, the identifications of shells by eye are inevitably subjective. To prioritize more "reliable" shells in our analyses, we first ranked the 11 identified shells by their appearance into three classes: "⊚" (undoubtedly expanding shells), " " (possibly expanding shells), and "△" (could be expanding shells). The shells in the "⊚"class exhibited clear ellipsoidal structures in the l-b-V space. The shells in the " "-class exhibited elliptical shapes in the l-b, l-V, or b-V plane. The sets of possibly fragmented ellipses are categorized into the "△"-class. Note that the physical parameters of the lower-ranked shells generally have larger uncertainties.  Table 1 for details on the identified shells.

Physical Parameters
Here, we estimate the physical parameters of the identified shells. The mass M of each shell can be derived from the sum of the column density of molecular gas associated to the shell. We assume that the entire molecular gas within each l-b-V ellipsoid, which are listed in Table 1 (Frerking et al. 1982), we calculated the column densities from the 12 CO J=1-0 line intensity.
The radius of each shell is defined as R ≡ √ ∆l ∆b/2. Assuming that each shell expands at constant velocity V exp , we can calculate the expansion time by t exp = R/V exp . The kinetic energy was calculated by E kin = M V exp 2 /2, and the kinetic power was calculated by P kin = E kin /t exp . The physical parameters M , E kin , t exp , and P kin , are also listed in Table 1.
The kinetic energy lies in the range of E kin = 10 49.5-51.8 erg. Because an SN explosion releases baryonic energy of (1-3) × 10 50 erg into interstellar space (Sashida et al. 2013), the kinetic energy of each ⊚-class shell corresponds to more than one SN. The expansion time is in the range of t exp = (5.6-16.6) × 10 4 yr. Unlike the shells in the l = −1. • 2 region (Tsujimoto et al. 2018), no t exp gradient along the spatial coordinates was observed. Except for two △-class shells, the calculated t exp coincide within a factor of 2, and V sys are confined in the range of 40 km s −1 width. The total kinetic en-    39.5 * • , , and △ are grading symbols widely used in Japan, which represent high, medium, and low evaluations, respectively. † "A"-"C" are the denotation in Tanaka et al. (2007). The "minor" and "major" shells were first noticed by Oka et al. (2001) ergy and total kinetic power of these shells amount to 10 51.9 erg and 10 39.5 erg s −1 , respectively.

Expanding-Shell Kinematics
Here, we examine the expanding-shell kinematics for the three ⊚-class shells (S2, N5, and N3) by the s-V plot method described in Sashida et al. (2013). In this method, accurate values of V sys and V exp are determined by fitting the uniform expansion model: to the molecular line data. s (x, y) is the normalized projected distance from the assumed center defined by where a and b are the assumed semimajor and semiminor axes, respectively, of the shell in the plane of the sky. Coordinates x and y correspond to those along the major and minor axes of the shell, respectively. In this study, we assumed that the major axes of the shells are parallel to the Galactic longitude or latitude. We adopted the larger ∆l/2 as a, and the smaller ∆b/2 as b.
We used the 12 CO J=3-2 data in this analysis. Before calculating the s-V plot, we applied the unsharp masking technique to the data to emphasize the ellipsoidal morphology of each shell in the l-b-V space. The spatial smoothing width was set to 0. • 05, and the velocity smoothing width to 25 km s −1 . The center positions as well as the semimajor and semiminor axes of the shells were carefully chosen by inspecting the CO J=3-2 and SiO J=2-1 data cubes. Then, we calculated s using Equation (3), to obtain the s-V plots shown in Figure  5.
Inner cavities can be observed in all s-V plots for ⊚class shells. The ellipse in the N3 plot can be clearly observed. The cavity in the N5 plot also exhibits a clear ellipse. These observations strongly indicate the expanding-shell kinematics. By contrast, the S2 plot exhibits a distorted ellipse, and the contribution of a clump at V LSR ≃ 60 km s −1 is not clear. The distortion could be due to deceleration caused by a dense clump in the northeast, or due to the overlapping of multiple expanding shells. The small clump in the S2 cavity at V LSR ≃ 120 km s −1 may not be physically related to the S2 shell.
By fitting Equation (1) in the range of s ≤ 1 to these s-V plots, we obtained V sys and V exp of S2, N5, and N3, resulting in V sys = 119.9, 142.0, and 144.0 km s −1 and V exp = 71.8, 50.0, and 44.0 km s −1 , respectively (Table   1). The N5 and N3 shells have similar V sys and V exp , while the S2 shell has a lower V sys and larger V exp . Despite the slight discrepancy in V sys , the s-V plot analyses clearly indicate the expanding-shell kinematics for the ⊚-class shells.
The s-V plots for and △-classes (lower two panels in Figure 5) also show elliptical cavities with velocity extents similar to V exp at V sys that are listed in Table 1. However, the s-extents of those cavities are less than 0.5, while those of ⊚-class shells reach ∼ 1.
These results demonstrate the difficulty in quantifying expanding-shell kinematics for lower-class shells.

SiO Line Emission
SiO is a well-established shocked gas tracer (e.g., Martín-Pintado et al. 1992). In the CMZ, where the molecular clouds exhibit highly turbulent kinematics, the SiO J=2-1 line is widely detected (e.g., Tsuboi et al. 2015;Takekawa et al. in preparation). Nevertheless, we expect some enhancement in SiO abundance just after the passage of dissociative shock. Because the expansion times of the shells in the l = +1. • 3 region is 10 4.7-5.2 yr, its current chemical properties must have been influenced significantly by shocks.
In Figure 3, green or cyan dominates the maps in velocities higher than 117 km s −1 , where the ellipses of S2 and N5 are evident. Cyan indicates that CO J=3-2 and SiO J=2-1 emissions are present, favoring the edges of the identified shells. This indicates a higher density and/or higher SiO abundance at the edges of the shells.
Here, we refer to the SiO J=2-1/H 13 CN J=1-0 (R SiO/H 13 CN ) and SiO J=2-1/H 13 CO + J=1-0 (R SiO/H 13 CO + ) intensity ratios to examine the enhancement of SiO abundance. These ratios are considered as indicators of shock strength (e.g., Handa et al. 2006). As the critical densities of these lines are similar (n cr ∼ 10 5-6 cm −3 ), they trace roughly the same spatial regions. Before the analyses, all datasets were smoothed with a 36 ′′ × 36 ′′ × 5 km s −1 full width at half maximum Gaussian function. Figure 6 shows the frequency histograms of R SiO/H 13 CN and R SiO/H 13 CO + weighted by the SiO J=2-1 intensity. The R SiO/H 13 CN distribution of the l = +1. • 3 region is similar to that of the CMZ, but slightly shifted to a higher ratio. However, the R SiO/H 13 CO + distributions are different. The l = +1. • 3 region has a larger proportion of gas with higher R SiO/H 13 CO + than the CMZ. The average values of R SiO/H 13 CN and R SiO/H 13 CO + for the l = +1. • 3 region are 1.47 and 7.52, both of which are larger than those for the CMZ (1.17 and 5.60, respectively). We performed the reduced χ 2 test to examine the equality of ratio distributions between CMZ and S2 N3 N5 l = +1. • 3 regions. The equality was rejected at a level of < 10 −16 , either for R SiO/H 13 CN and R SiO/H 13 CO + distributions, if we employ the typical intensity reproducibility (8%) as uncertainties in the SiO intensity. These results suggest that the SiO abundance is enhanced in the l = +1. • 3 region.
The spatial and velocity distributions of high R SiO/H 13 CN gas are shown in Figure 7. The threshold was set to R SiO/H 13 CN = 1.5, which is roughly the average value for the l = +1. • 3 region. High R SiO/H 13 CN gas is concentrated at the locations of shells N2-N5, B, and Arc. The most prominent concentration is found at the southwestern edge of N5. A filament of high R SiO/H 13 CN gas is observed along the eastern edges of S2, N7, and N6. In the latitude-velocity map, high R SiO/H 13 CN gas favors high intensity areas at velocities around V LSR = 100 km s −1 . In other words, the distributions of high R SiO/H 13 CN gas do not highlight the identified expanding shells. It roughly traces the distribution of high-density gas. This means that SiO abundance is not particularly enhanced at the identified expanding shells, while it is enhanced in the entire l = +1. • 3 region.
An explanation as to why the SiO abundance is not enhanced in the expanding shells with rather short (∼ 10 5 yr) expansion times may be that the abundance becomes saturated in the l = +1. • 3 region. The SiO fractional abundance in the "SiO clouds" in the CMZ is ∼ 10 −9 (Martín-Pintado et al. 1997), which is similar to those in regions that experienced fast shocks. As the frequent The broad-velocity-width nature, association of the shells/arcs and molecular flares, definite evidence for expanding kinematics in the shells, and abundance enhancement of a shock-origin molecule toward the region all support the notion that the l = +1. • 3 region has been accelerated by multiple SN explosions. Such a region is called a "molecular superbubble" (e.g., Oka et al. 2001;Tanaka et al. 2007;Tsujimoto et al. 2018). Such a superbubble must contain a young star cluster that yields SNe frequently. The absence of H II regions indicate a cluster age of ≥ 10 Myr, the main-sequence lifetime of a 15 M ⊙ star. The presence of type-II SNe indicates that the cluster is younger than 30 Myr. This cluster age implies that the l = +1. • 3 region may have experienced SNe over several Myr, which is consistent with the vertically elongated structure and widespread high R SiO/H 13 CN over the region.
The total kinetic power of the l = +1. • 3 region is P kin ≃ 10 39.5 erg s −1 . As an SN explosion releases (1-3) × 10 50 erg energy into interstellar space (Sashida et al. 2013), this kinetic power corresponds to an SN rate of ∼ 10 −3.1 -10 −3.5 yr −1 . Assuming that the star cluster formed instantaneously, i.e., the stellar members have the same age, the cluster mass is estimated to be M cl ∼ 10 7.5 M ⊙ , employing the Scalo initial mass function (IMF; Scalo 1986) with the low-/high-mass cutoffs and mass of the heaviest stars as 0.08 M ⊙ /100 M ⊙ and 8 M ⊙ , respectively. This unbelievably high cluster mass is a challenge that is yet to be overcome in the molecular superbubble hypothesis for the l = +1. • 3 region.
The absence of a bright infrared counterpart presents another challenge to the hypothesis. A massive cluster with an initial mass of M cl ∼ 10 7.5 M ⊙ and age of 10-30 Myr should have a total luminosity of L cl ∼ 10 9.0 L ⊙ (Williams & Perry 1994). However, the far infrared (100 µm) luminosity of the l = +1. • 3 region is only L IR ∼ 10 6.4 L ⊙ , which is 2.6 orders of magnitude lower than the estimated total luminosity. This situation is similar to that encountered at the l = −1. • 2 region (Tsujimoto et al. 2018). This serious discrepancy between L cl and L IR could be partly explained by an abnormal IMF with a shallower slope and/or higher lowmass cutoff for the putative star cluster. The other types of explosions, e.g., Type Ia SNe and neutron star mergers (Rosswog et al. 2013) can also contribute to the kinetic power. Some extreme ideas such as the dark stellar remnant clusters should be considered in future studies.
1. We obtained high-quality maps of multiple molecular lines, including high-density and shocked gas probes in the l = +1. • 3 region.
2. The l = +1. • 3 region exhibits higher R SiO/H 13 CN and R SiO/H 13 CO + than those in the CMZ. This indicates that a strong shock has passed through this region, enhancing the SiO emission.
4. We ranked the shells into three classes based on how apparent their elliptical shapes are in the l-b-V space, and confirmed clear expanding motion for three shells in the highest ⊚-class. These results support the hypothesis that the l = +1. • 3 region is a superbubble.
5. The total kinetic energy of the shells are E kin ∼ 10 52 erg, and their expansion times are typically t exp ∼ 10 5 yr. Then, the SN rate of this region is estimated to be ∼ 10 −3.1 -10 −3.5 yr −1 .
6. For the molecular superbubble hypothesis, there needs to be young (10-30 Myr) massive (M cl ∼ 10 7.5 M ⊙ ) cluster. The kinematics and morphologies of the molecular gas and energetics of this region generally support this scenario.
These new results reinforced the molecular superbubble hypothesis for the l = +1. • 3 region, while the sharp discrepancy between the far infrared and theoretical luminosities of the embedded cluster challenges the scenario. This work also demonstrated the importance of broad-velocity-width compact molecular features in searching for localized energy sources hidden behind severe interstellar extinction and stellar contamination.