SIGNATURES OF YOUNG STAR FORMATION ACTIVITY WITHIN TWO PARSECS OF Sgr A*

2.1.1. H₂O Line at 22 GHz

2.2.1. SiO (2-1) and (5-4) Lines

2.2.2. [Ne ii] Line and 12.8 μm Continuum

2.2.3. Mid-IR Data

We detect 13 water masers concentrated toward the inner edge of the molecular ring. Figure 1 shows the distribution of water masers (crosses) superimposed on an image of the velocity-integrated HCN (1-0) line (Christopher et al. 2005); The insets show their spectra. Entries in Columns 1–7 in Table 1 give Galactic coordinates, source name, celestial coordinates, the peak flux, the peak velocity, and cross references for all detected masers. A large-scale water maser survey toward the Galactic center using ATCA did not detect any of the sources reported here because of their lower sensitivity with rms noise σ ∼ 0.1 Jy (Caswell et al. 2011). Table 1 lists multiple velocity components in several sources with peak velocities ranging between ∼−91 and 67 km s⁻¹. The peak fluxes range between 22.6 and 319.6 mJy beam⁻¹ , corresponding to brightness temperatures ranging between ∼0.6 and 7.9 K. These sources are spatially unresolved, and thus these brightness temperatures are lower limits. We regard these sources as masers because of their narrow line widths of ∼1 km s⁻¹.

Zoom In Zoom Out Reset image size

The HCN (1-0) observations resolved 26 cores with a typical diameter of ∼7'' (0.25 pc). Three collisionally excited methanol masers coincide with 3 of 26 HCN clumps F, G, and V, drawn as circles centered on masers in Figure 1 (YBWR08). The region to the NE of the molecular ring in Figure 1 shows three water masers (1, 11, and 12) and two methanol masers coincident with the clumps labeled F and G, in the HCN (1-0) line map (Christopher et al. 2005; YBWR08). The methanol spectrum in the vicinity of clump F has three narrow velocity peaks between 52 and 55 km s⁻¹ and a broad redshifted wing extending up to 100 km s⁻¹ (YBWR08). This region is inferred to be a star-forming site because of the collisionally excited methanol maser at 44 GHz and the redshifted broad wing in the HCN (1-0) line with a peak velocity of ∼60 km s⁻¹ (YBWR08). The velocity correlation of 22 GHz masers and molecular line emission, SiO (5-4) and HCN (1-0), will be discussed in Section 3.3.3.

It is possible that interstellar water lines are contaminated by water lines from evolved OH/IR stars in this highly confused region of the Galaxy. To exclude this possibility, we compared the position of new water masers listed in Table 1 with those of OH/IR stars and SiO masers associated with evolved stars. With the exception of sources 1 and 13 (Lindqvist et al. 1992; Levine et al. 1995; Sjouwerman & van Langevelde 1996; Sjouwerman et al. 2002; Li et al. 2010), we did not find evolved stellar counterparts with similar velocities and positions to those of detected water masers in Table 1. Water maser 1 coincides with an OH/IR star 359.956-0.050 2a (Table 2 of Sjouwerman et al. 2002), although it is not clear why this water maser has four velocity components 47, 54, 58, and 63 km s⁻¹ (Table 1) when the velocity profile of the OH/IR star has only two velocity components (Sjouwerman et al. 2002). These velocity components match up with those of collisionally excited methanol line emission (YBWR08). Similarly, maser 13 also shows five velocity peaks, 78.53, 80.22, 81.90, 87.81, 91.and 19 km s⁻¹ (Table 1),  and is spatially coincident with the OH/IR star 359.970-0.049, with two velocity peaks of 37.5 and 94.5 km s⁻¹ (Sjouwerman et al. 2002). The spectrum of maser 8 appears to be similar to that of an OH/IR star.

Among other water masers, we note that water maser 11 shows two forbidden velocities −63 and −71 km s⁻¹ assuming orbits aligned with the Galactic rotation. An SiO maser source, denoted SiO-18 in Table 2 of Li et al. (2010) is located within a few arcseconds of water maser source 4. However, the peak emission from SiO-18 is at 39.6 km s⁻¹ whereas water maser source 4 has a peak velocity of −82 km s⁻¹, suggesting that they are not associated with each other. We note that water maser 6 has more than three peaks like water maser 1 and 13. OH/IR stars are identified by two peaks symmetrically centered around the velocity of the central star, so water masers with more than two velocity components are unlikely to be associated with individual OH/IR stars. It is possible that these sources with multiple velocity components are interstellar water masers but are contaminated by OH/IR stars. Water maser 8 has two components that are much more widely separated in velocity than other water masers' components. In summary, we have detected a total of five water masers with three or more (1, 5, 6, 11, and 13) velocity components, five water masers with a single velocity component (3, 4, 9, 10, and 12) and three water masers with two velocity components (3, 7, and 8). Water masers with double velocity components are probably associated with OH/IR stars whereas the remaining sources are likely to be associated with interstellar masers.

The detection of water and collisionally excited methanol masers motivated us to search for YSOs within the inner parsec of Sgr A*. This region is populated by about 60 dusty and red sources that could be OH/IR stars and/or YSO candidates, both of which have strong infrared excess emission (Viehmann et al. 2005, 2006). OH/IR stars represent evolved AGB stars at the end of their lives (Habing 1996) whereas YSOs are characterized by dusty envelopes and disks that absorb the radiation from the central protostar (e.g., Whitney et al. 2003). Near- and mid-IR photometric observations with VLT have identified a total of 64 sources in the central parsec at wavelengths ranging between 1.6 μm (H band) and 19.5 μm (Q band), respectively (Viehmann et al. 2006). Four types of sources were classified based on their SEDs (Viehmann et al. 2006; Preger et al. 2008). Type I consists of luminous IR sources that are embedded in the N arm ionized streamers. These sources (e.g., IRS 1, IRS 2L, IRS 10W, and IRS 21), showing bow-shock structures with featureless SEDs that peak in the mid-IR wavelengths, are generally thought to be interacting with the mini-spiral (Tanner et al. 2005). Type II consists of low-luminosity sources that have similar bow shock morphology to those of Type I sources, but are offset from the mini-spiral streamers. Type III sources are cool stars with SEDs that peak in the near-IR wavelengths. Last, Type IV sources are hot stars (e.g., IRS 16NE, IRS 16NW, AF, and AHH) with SEDs that peak at shorter wavelengths. There are also a number of sources that could not be classified among these four types based on their SEDs (Viehmann et al. 2006).

We carried out YSO SED fitting of dusty sources of all types listed in Table 1 of Viehmann et al. (2006) to provide additional constraints on the nature of these sources. We used the same technique that identified a population of YSO candidates toward the inner few 100 pc of the Galaxy (Yusef-Zadeh et al. 2009, but see Koepferl et al. 2015). To obtain reliable fits, we only selected sources with at least four flux measurements (52 out of 64 sources). The Viehmann et al. (2006) observations used nine VLT/VISIR bands in the wavelength range from 1.6 to 19.5 μm, with an angular resolution (FWHM) of 03–06. We considered using the catalog of near-IR sources (Schödel et al. 2010) for additional data points for those sources that had limited data in Viehmann et al. (2006). However, the lack of proper motion data and confusing sources prevented us from using the new catalog at H, J, and L' bands (Schödel et al. 2010). The availability of data from a single instrument that provides a broad wavelength range and high angular resolution minimizes source confusion and systematic effects.

We compare the sources' SEDs to the Robitaille et al. (2006) large grid of pre-computed YSO model SEDs using the linear regression SED fitting tool developed by Robitaille et al. (2007). We consider models with interstellar extinction (A_V) between 25 and 30 mag, and distance between 7.65 and 9.35 kpc (i.e., the distance to the Galactic center of 8.5 kpc ± 10%). We identify well-fit models as those with a normalized ${\chi }^{2}$ per data point (${\chi }^{2}/{pt}$) less than 3. As a result, only 19 out of 52 sources are well-fit by YSO models. A good YSO model fit indicates that the source's SED is consistent with the YSO model SED, but does not provide a definite classification.

We estimate physical parameters of these sources by averaging parameters of all YSO models that fit the source's SED with ${\chi }^{2}/{pt}$ in a range between the ${\chi }^{2}/{pt}$ for the best-fitting model (${\chi }_{\mathrm{min}}^{2}/{pt}$) and (${\chi }_{\mathrm{min}}^{2}/{pt}+2$). The visual examination of the fits showed that the YSO model fits with ${\chi }^{2}/{pt}$ within this arbitrary range provide acceptable fits to the SEDs (see Figure 2). The fitting results are presented in Table 2. In Column 1, we provide the Viehmann et al. (2006) ID numbers (V06 ID). Columns 2 and 3 list the ${\chi }^{2}/{pt}$ for the best-fit model (${\chi }_{\mathrm{min}}^{2}/$) and the number of acceptable fits (n_fits), respectively. In Columns 4–15, we provide the estimates of the physical parameters for the best-fit model, as well as the average physical parameters calculated based on all the acceptable fits (see above). The physical parameters include the interstellar extinction (A_V), stellar luminosity (L ${}_{\star }$), temperature (T ${}_{\star }$), and mass (M ${}_{\star }$), as well as the envelope mass (M_env), and disk mass (M_disk). The best and average evolutionary stages (Columns 16–17) are based on sources' physical parameters as defined by Robitaille et al. (2006). Column 18 gives the number of data points used for the SED fitting.

Zoom In Zoom Out Reset image size

Table 2.  The SED Fitting Results: Physical Parameters of the YSO Candidates

V06	${\chi }_{\mathrm{min}}^{2}/$	n ${}_{\mathrm{fits}}$ b	AV		L ${}_{\star }$ (104 ${L}_{\odot }$)		T ${}_{\star }$ (104 K)		M ${}_{\star }$(${M}_{\odot }$)		Menv(${M}_{\odot }$)		Mdisk(${M}_{\odot }$)		Evol. Stage		#
V0 ID	pt a		Best	Avec	Best	Ave	Best	Ave	Best	Ave	Best	Ave	Best	Ave	Best	Ave	Data
5	1.36	367	25.0	25.3 (1.0)	2.4	1.1 (0.05)	3.21	2.74 (0.02)	15.9	11.7 (0.2)	0.00	0.15 (0.07)	0.000	0.006 (0.001)	III	II	6
6	1.33	31	25.0	25.0 (0.0)	7.1	6.0 (0.27)	3.76	3.66 (0.03)	23.7	22.1 (0.4)	0.00	0.0 (0.0)	0.043	0.130 (0.027)	II	II	9
10	0.53	5700	25.5	27.2 (2.4)	0.12	0.33 (0.01)	1.93	1.94 (0.01)	6.4	7.85 (0.03)	9.2	22.1 (1.8)	0.056	0.033 (0.001)	II	I	4
11	0.03	4583	29.8	27.7 (2.4)	0.32	0.37 (0.01)	0.51	1.88 (0.01)	10.6	8.55 (0.03)	1293	52.3 (3.3)	0.219	0.054 (0.002)	I	I	5
12	0.03	3039	29.2	28.3 (2.2)	0.49	0.46 (0.01)	0.41	1.98 (0.01)	11.9	9.21 (0.04)	1415	65.7 (4.8)	0.000	0.061 (0.002)	I	I	5
13	0.43	29	25.0	25.1 (0.6)	7.09	6.58 (0.51)	3.77	3.45 (0.05)	23.7	22.8 (0.8)	0.00	160 (114)	0.043	0.137 (0.028)	II	I	7
14	0.27	627	29.1	26.9 (2.2)	2.43	1.77 (0.04)	3.21	2.98 (0.01)	15.9	13.8 (0.1)	0.00	2.7 (1.5)	0.010	0.029 (0.003)	II	I	7
15	0.19	2192	29.9	26.8(2.2)	1.15	0.87 (0.01)	2.84	2.43 (0.01)	12.2	11.2 (0.1)	0.00	83.0 (7.2)	0.004	0.060 (0.003)	II	I	7
18	0.49	1245	25.0	26.7 (2.2)	0.94	1.28 (0.02)	2.75	2.75 (0.01)	11.4	12.2 (0.1)	0.00	21.8 (4.9)	0.066	0.042 (0.003)	II	I	7
19	0.06	5145	25.0	27.0 (2.3)	1.17	0.93 (0.01)	2.85	2.41 (0.01)	12.3	11.39 (0.03)	20.4	81.5 (3.2)	0.000	0.067 (0.002)	II	I	7
26	0.24	4954	25.0	27.6 (2.4)	0.21	0.17 (0.003)	0.43	1.66 (0.01)	10.1	6.40 (0.03)	33.7	13.8 (1.5)	0.040	0.026 (0.001)	I	I	5
28	1.85	776	25.0	25.1 (0.6)	1.21	1.61 (0.03)	2.81	2.92 (0.01)	12.0	13.2 (0.1)	0.00	0.0 (0.0)	0.038	0.016 (0.002)	II	II	8
31	1.98	193	25.9	26.3 (2.0)	2.76	3.16 (0.08)	3.27	3.31 (0.01)	16.6	17.2 (0.2)	0.00	0.0 (0.0)	0.000	0.034 (0.005)	II	II	7
32	2.37	547	25.0	25.3 (1.0)	2.68	1.96 (0.03)	3.25	3.00 (0.02)	16.4	14.6 (0.1)	0.00	24.4 (3.7)	0.018	0.037 (0.003)	II	I	9
39	0.80	879	30.0	27.8 (2.3)	1.62	1.33 (0.03)	0.41	2.57 (0.03)	23.9	12.7 (0.1)	408	87.8 (13.1)	0.221	0.056 (0.004)	I	I	5
47	1.32	186	30.0	27.8 (2.0)	4.62	2.84 (0.09)	3.52	3.23 (0.02)	20.0	16.4 (0.2)	0.00	0.0 (0.0)	0.068	0.016 (0.003)	II	II	9
48	0.36	565	30.0	27.5 (2.2)	3.86	1.98 (0.04)	3.39	3.04 (0.01)	18.1	14.3 (0.1)	0.00	0.7 (0.7)	0.052	0.027 (0.003)	II	II	8
54	0.48	19	25.0	25.0 (0.0)	7.09	6.63 (0.37)	3.76	3.70 (0.04)	23.7	22.9 (0.6)	0.00	0.0 (0.0)	0.043	0.176 (0.030)	II	II	9
57	0.77	2549	30.0	27.3 (2.2)	0.40	0.55 (0.01)	2.38	2.35 (0.082)	8.8	9.19 (0.04)	0.00	14.6 (1.8)	0.007	0.027 (0.001)	II	I	6

Notes.

^a ${\chi }_{\mathrm{min}}^{2}/{pt}$ is a ${\chi }^{2}$ per data point for the best-fit model. ^b n ${}_{\mathrm{fits}}$ is a number of fits with ${\chi }^{2}/{pt}$ between ${\chi }_{\mathrm{min}}^{2}/{pt}$ and ${\chi }_{\mathrm{min}}^{2}/{pt}+2$. ^cThe uncertainties represent standard deviations of the mean.

Download table as:  ASCIITypeset image

Stage I objects are young protostars embedded in an opaque infalling envelope. In Stage II are objects the envelope has mostly dispersed, and the central star is surrounded by an opaque disk. In Stage III the disk is optically thin. The SEDs of sources that are well-fit with YSO models are shown in Figure 2. In each plot, the best-fit model and all acceptable models are shown in black and gray, respectively. The dashed line is the central stellar atmosphere corresponding to the best-fit model, extincted by the fitted foreground extinction. To place additional constraints on best-fit models, we explored the possibility of using Herschel data at long wavelengths. However, the Herschel PACS instrument (Poglitsch et al. 2010) had insufficient angular resolution (∼55) at 70 μm to resolve individual YSO candidates in this crowded field and in the presence of diffuse emission from the CMR. Despite this, for some sources, the generous upper limits that are derived, ranging between (3.24 and 5.75)$\;\times \;{10}^{-9}$ erg s⁻¹ cm⁻² for sources listed in Table 2, can rule out a few of the models with the most massive envelopes. Somewhat more restrictive upper limits (e.g., factors of 10 lower) would eliminate models with envelopes greater than ∼14 ${M}_{\odot }$ . Sources 10, 11, 26, and 39 would be most strongly impacted, changing them from likely Stage I objects to Stage II objects.

SED modeling of the infrared excess sources listed in Table 2 indicates that they are massive YSO candidates. In particular, the SED fits to members of the two complexes, IRS 5 and IRS 13N, are consistent with being massive YSO candidates. Four members of the IRS 5 Complex (IRS 5, IRS 5E, IRS 5S, and IRS 5SE1), as well as sources IRS 13N, IRS 13NE, IRS 2L, IRS 1W, and IRS 10W are classified as main-sequence or WR stars that may be associated with ionized streamers. Stellar sources in the IRS 5 Complex have been interpreted as low-luminosity bow shock sources, low-luminosity dust forming AGB stars, or YSOs (Preger et al. 2008). Based on proper motion measurements of cluster members in IRS 5 and IRS 13 as well as their high luminosities, these sources are unlikely to be AGBs and could be members of young cluster of stars (Preger et al. 2008; Eckart et al. 2013). The distribution of luminous YSO candidates imply that ongoing star formation is taking place near Sgr A* within the ionized streamers orbiting Sgr A* (e.g., Viehmann et al. 2006; YBWR08; Eckart et al. 2013; Nishiyama & Schödel 2013; Yusef-Zadeh et al. 2013, 2014; Jalali et al. 2014). Additional support for ongoing star formation within 2 pc of Sgr A* comes from radio continuum and near-IR polarization studies, as described below. Radio continuum observations did not detect a bow shock structure but resolved IRS 21 into at least four components (Yusef-Zadeh et al. 2014), suggesting that this a compact young stellar cluster. The radio emission from a cluster of four YSOs is due to thermal bremsstrahlung from ionized gas that is being photo-evaporated from a disk by the strong UV radiation field at the Galactic center.

In our sample of YSO candidates we include two sources that were not well fitted with the YSO models. These are IRS 1W and IRS 10W with ${\chi }_{\mathrm{min}}^{2}/$ of 12 and 5, respectively. Their SEDs are shown separately in Figure 3. These sources are not included in Table 2 since the physical parameters estimated from the fitting are highly uncertain. However, other observations indicate that these sources are likely YSO candidates. IRS 1W and IRS 10W coincide within three clumps of SiO (5-4) emission 1, 2, and 3 (Yusef-Zadeh et al. 2013). SiO (5-4) line emission with broad line widths, as discussed in Section 3, provide strong support for protostellar outflows interacting with the ISM. Additional support comes from near-IR polarization measurements indicating that sources IRS 1W, IRS 10W, and IRS 21 are intrinsically polarized sources (Yoshikawa et al. 2013). Intrinsically polarized sources are argued to be associated with YSOs driven by outflows in star-forming regions (Yoshikawa et al. 2013). Table 3 gives the Viehmann et al. (2006) ID numbers (V06 ID), the corresponding near-IR names (Names), and equatorial coordinates for 19 well-fitted YSO candidates, as well as sources IRS 1W and IRS 10W. Cross-references are provided for some of the sources.

Zoom In Zoom Out Reset image size

Recent radio continuum observations within 30'' of Sgr A* show evidence of partially resolved free–free emission from near-IR identified stellar sources, e.g., IRS 21, IRS 5, IRS 13N, and IRS 13E (Yusef-Zadeh et al. 2014). We argued that the radio emission is due to thermal bremsstrahlung from ionized gas that is being photo-evaporated from a disk by the strong UV radiation field at the Galactic center. In this picture, the mass-loss rate due to photo-evaporation implies the existence of a reservoir of neutral material. Robitaille et al. (2006) models consider envelope outer radii ranging between $\sim 1.5\times {10}^{16}$ and $\sim 1.5\times {10}^{18}$ cm. This range in size appears to be motivated by the limit at which the temperature would drop to 30 K. On the other hand, disk and inner envelope radii from our SED models range between 5 and 3000 AU (7.5$\;\times \;{10}^{13}$ and $4.5\times {10}^{16}$ cm). In the case of evolved stars in the Galactic center region, there is no identified AGB star that is detected at radio wavelengths. The outer radii of AGB stars can be as large as few $\times {10}^{17}$ cm (Wallerstein & Knapp 1998). If the envelope of AGBs or YSOs were photoionized, their angular size at the Galactic center would show extended radio sources greater than 1'' (1'' is 0.04 pc at the distance of the Galactic center). One of the dustiest AGB sources in the Galactic center is IRS 3 (Preger et al. 2008), which shows no signature of extended thermal emission. Radio sources are partially resolved at ∼01 resolution. Thus, detection of partially resolved radio continuum sources in the Galactic center is consistent with SED models of infrared excess sources. Notable sources indicating a disk of material associated with YSO candidates are IRS 1W, IRS 5, IRS 5NE, IRS 13N, IRS 13NE, IRS 21, and IRS 34SW, corresponding to sources 1, 13, 14, 47, 48, 54, and 57 in Table 2.

Some of the stars, IRS 1W, IRS 10W, IRS 5, IRS 2L, and IRS 21, show bow shock structures with featureless SEDs that peak in mid-IR wavelengths (Tanner et al. 2002, 2005; Clenet et al. 2004; Muzic et al. 2010; Sanchez-Bermudez et al. 2014). The bow shock morphologies of these sources are interpreted in terms of the interaction of ionized winds with the streamers of ionized gas (Tanner et al. 2005; Sanchez-Bermudez et al. 2014). This model adopts a spherical outflow from central Wolf–Rayet stars (Tanner et al. 2002, 2005). Alternatively, the interaction of ionized streamers could take place with the material evaporating from the YSO disks assuming that the outflow from the YSO disks is faster than the relative motion between the YSOs and the ionized streamers. In this scenario, the cause of the asymmetry comes from the outflow arising preferentially from the disk surface, facing the ionizing sources in the central cluster of stars. The asymmetry in the bow shock geometry could also come from YSO disks interacting with ionized streamers. The orientation of YSO disks should be random in the disk picture. The orientations of the bow shock in IRS 10W, IRS 5 do not seem to follow the north–south direction of the streamers, which is consistent with disks associated with YSO candidates.

YSO candidates' stellar masses derived from the SED fitting can be used to estimate the star formation rate (SFR; e.g., Whitney et al. 2008; Yusef-Zadeh et al. 2009; Povich et al. 2011). The SFR is calculated by estimating the total mass of the YSO candidates and dividing it by the YSOs' approximate lifetime. Since our sample of YSO candidates is small, we can only provide a rough approximation of the SFR. We use Stage I (the average evolutionary stage from Table 2) YSO candidates only to get a sample of sources at approximately the same age and lifetime.

To estimate the total mass of the YSOs, we construct the stellar mass function, scale the Kroupa (2001) initial mass function (IMF; ${dN}/{dm}\propto {m}^{-\alpha }$, where m is a stellar mass) to match the peak of the distribution, and integrate under the IMF over a mass range of 0.08–50 ${M}_{\odot }$. The Kroupa (2001) IMF has a slope (α) of 1.3 between 0.08 and 0.5 ${M}_{\odot }$ and 2.3 between 0.5 ${M}_{\odot }$ and 50 ${M}_{\odot }$. The estimated total mass of the YSO candidates is ∼1318 ${M}_{\odot }$. Assuming a typical lifetime of the Stage I YSO of 10⁵ yr, the SFR within 1 pc of Sgr A* is estimated to be ∼1$\times {10}^{-2}$ ${M}_{\odot }$ yr⁻¹. This is only an order magnitude estimate because of the small number statistics, uncertainty of a factor of two in the age of Stage I YSOs (see Koepferl et al. 2015), and the extrapolation of the IMF down to the cut-off at 0.08 ${M}_{\odot }$.

3.2.1. YSO Candidates in Rotating Disks Orbiting Sgr A*

One of the most intriguing aspects of the YSO candidates is their spatial distribution, which appears to be similar to that of the young, hot stars orbiting Sgr A*. The so-called edge-on clockwise and face-on counterclockwise stellar disks have positive and negative angular momentum j, respectively (Beloborodov et al. 2006; Paumard et al. 2006) and consist of hot and massive stars which are a few million years old (Paumard et al. 2006; Lu et al. 2009). The two disks extend out to 10'' from Sgr A*. The position angles (PAs) of young stars in the clockwise and counterclockwise stellar disks are ∼30° and 130°, respectively. Figures 4(a) and (b) show the positions of YSO candidates superimposed on a 34 GHz continuum and 3.8 μm images, respectively. YSO candidates IRS 5, source 28, and IRS 2L are embedded or offset to the east of the N arm of the ionized mini–spiral, IRS 4 and source 12 lie in the E arm, and IRS 13E, IRS 13N, and IRS 34SW lie in projection against the bar of ionized gas and its NW extension. The YSO candidates that are distributed diagonally from NE to SW run parallel to the face-on counterclockwise disk whereas the sources to the SE fall along the clockwise stellar disk. The distributions of YSOs extend to ∼15'' from Sgr A*, larger than the 10'' extent of counter-rotating disks. The three-dimensional motions of many of the YSO candidates are unknown; however, the line of sight motion of the YSO candidates and their apparent distribution on the sky show a trend that is consistent with the positive and negative angular momenta of two rotating disks. For example, the proper motion of the IRS 5S1 to 3 clusters suggest that the overall angular momenta of these sources is in agreement with the clockwise rotating disk of stars (Preger et al. 2008). In addition, IRS 34SW has positive angular momentum, and thus follows clockwise rotating disks whereas IRS 13 has negative angular momentum values. This suggests that the YSO candidates are forming in the residual outer regions of the original gas disks that formed the original stellar disks orbiting Sgr A*. Future proper motion measurements will determine the angular momentum of the YSO candidates and test the picture proposed here.

Zoom In Zoom Out Reset image size

Another indication that the region within 0.5 pc of Sgr A* is a site of ongoing star formation comes from clumps of SiO emission found in this region (Yusef-Zadeh et al. 2013). ALMA observations of the interior of the molecular ring detected 11 clumps of SiO (5-4) emission within the inner 0.5 pc of Sgr A*. SiO line emission is generally associated with shocks from protostellar outflows, so SiO clumps detected within the molecular ring were interpreted as evidence of young star formation activity near Sgr A* where outflows from YSOs interact with the surrounds. The location of five SiO clumps are offset from the eastern edge of the N arm but run parallel along the N arm of ionized gas. These sources show typical line widths between 11 and 21 km s⁻¹ and peak radial velocities that decrease from 37.8 to 7.4 km s⁻¹ to the N of the streamers. This trend in radial velocity is consistent with the trend in the kinematics of ionized gas of the N arm (e.g., Zhao et al. 2010). However, the radial velocity distribution of the SiO clumps does not match the kinematics of ionized gas in the N arm. The positions of two of the most prominent SiO (5-4) clumps found by ALMA (sources 1 and 11 in Yusef-Zadeh et al. 2013) lie along the extension of YSO candidates found to the NE of where the IRS 5 cluster is located. This suggests that there is a concentration of molecular gas to the east of the N arm.

We extended the study of SiO (5-4) line emission from the inner 0.5 pc to the molecular ring within ∼2 pc of Sgr A* using data taken with the SMA with a resolution of 36 × 24 (Martin et al. 2012). Table 4 lists the parameters of Gaussian fitted SiO (5-4) line emission found in the molecular ring. Columns 1–7 show the source number, following the SiO (5-4) sources detected by ALMA (Yusef-Zadeh et al. 2013), celestial coordinates, peak intensity, FWHM, and full width at zero intensity (FWZI), respectively. The SiO (5-4) sources detected by ALMA are concentrated in a region within the molecular ring where SMA observations could not detect any SiO (5-4) line because of its poor sensitivity with the exception of source 10, which was detected in both ALMA and SMA observations. Table 4 shows that the parameters of the fit to this source detected with ALMA and SMA are similar to each other. The discrepancy between the fitted parameters of SMA source 10 and 10 (ALMA) in Table 4 is mainly due to different spatial and spectral resolutions and the sensitivity of the ALMA and SMA data.

Figure 5 shows the velocity-integrated SiO (5-4) line intensity of 26 clumps toward the molecular ring. To investigate if these new SiO clumps have similar characteristics to known protostellar outflows in the Galactic disk, Figure 6(a) compares the line widths and SiO (5-4) luminosity of the emission from 11 clumps interior to the ring (red), found from ALMA observations, 26 clumps (12–36) in the molecular ring listed in Table 4 (blue) as well as low- and high-mass protostars in the Galactic disk (black). The SiO sources lying in the central hole in the central molecular ring as well as the clumps coincident with the ring show a similar lack of dependence between the FWZI and SiO luminosity found for high-mass protostellar systems in the Galactic disk (Yusef-Zadeh et al. 2013). This supports the suggestion that the SiO clumps in the CMR are massive protostellar outflows and the CMR itself is a site of recent massive star formation. We note a bipolar outflow candidate (Figure 6(b)) in one of the SiO clumps, as described below.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

3.3.1. SiO (5-4) Bipolar Outflow

3.3.2. LVG Analysis of SiO Lines

In order to apply an LVG analysis to the sources detected in the SiO (5-4) and (2-1) lines, we first convolved the CARMA and SMA data to an identical resolution of 685 × 373 and then performed Gaussian fits to the sources, obtaining the parameters listed in Table 5. The CARMA sensitivity of SiO (2-1) was not sufficient to detect source 12. In addition, SiO (5-4) and SiO (2-1) data from SMA and CARMA showed some clumps with multiple velocity components. The LVG analysis adopted a uniform slab geometry for the emitting region, and included a radiation field characterized by a dust temperature of 75 K, 30 mag of visual extinction, and a dust extinction curve following Draine & Lee (1984). We computed the emergent intensities in the two lines given the slab's assumed gas density, temperature, FWHM velocity, and SiO column. In Figure 7, we show contours of 5-4 versus 2-1 brightness temperature for constant hydrogen number density (n_H) and for fixed SiO column density, assuming a FWHM of 30 km s⁻¹ for kinetic temperatures of 150 K. Also plotted are the points corresponding to the observed SiO sources, which lie close to and follow contours within ${n}_{{\rm{H}}}={10}^{6}$ or n(H2) $\sim \;5\times {10}^{5}$ and ${10}^{6.15}$ cm⁻³ for the assumed kinetic temperature of 150 K. These estimates are similar to those given by Amo-Baladron et al. (2009). Note that the predicted brightness temperatures scale linearly with SiO column density because the lines are optically thin. We conclude that the sources have ${n}_{{\rm{H}}}$ ≈ 1–4 × 10⁶ cm⁻³ for assumed kinetic temperatures of 50 and 150 K and N(SiO) ≈ 10^14.4 cm⁻². As with other linear molecules, we find that the density inferred from the LVG analysis is inversely proportional to the assumed kinetic temperature, so that the LVG constrains the pressure in the SiO sources to be $P/k\approx 1\times {10}^{8}\;{\rm{K}}\;{\mathrm{cm}}^{-3}$.

Zoom In Zoom Out Reset image size

Table 5.  Gaussian Line Parameters of Fitted SiO (5-4) and (2-1) Sources

	$\underline{\mathrm{CARMA}}$			$\underline{\mathrm{SMA}}$
	Peak	Zentrum	FWHM	Peak	Zentrum	FWHM
	(mJy beam−1)	(km s−1)	(km s−1)	(mJy beam−1)	(km s−1)	(km s−1)
10	45.2 ± 64.2	−40.8 ± 3.2	46.8 ± 7.6	167.2 ± 18.3	−41.8 ± 1.3	24.2 ± 3.0
12	...	...	...	218.8 ± 10.1	−9.4 ± 1.7	74.4 ± 4.0
13	87.3 ± 13.2	−19.2 ± 3.2	61.1 ± 11.0	353.2 ± 23.6	−18.2 ± 0.7	23.7 ± 1.8
14	116.0 ± 8.7	−38.4 ± 1.1	32.0 ± 2.7	352.9 ± 20.9	−35.8 ± 0.9	31.0 ± 2.1
15	131.7 ± 9.6	−50.8 ± 1.5	43.1 ± 3.6	488.9 ± 23.8	−49.0 ± 1.0	42.2 ± 2.3
16	162.0 ± 9.9	−60.0 ± 0.7	26.4 ± 1.8	461.8 ± 39.5	−58.4 ± 1.1	26.3 ± 2.6
17	20.2 ± 12.7	27.2 ± 0.6	21.4 ± 1.5	420.7 ± 44.6	30.5 ± 0.6	12.0 ± 1.4
18	230.9 ± 12.1	30.7 ± 0.4	16.4 ± 1.0	395.6 ± 20.9	30.7 ± 0.3	13.3 ± 0.8
19	244.7 ± 12.7	28.0 ± 0.2	10.4 ± 0.6	403.4 ± 46.2	28.5 ± 0.5	10.1 ± 1.3
20	83.6 ± 15.9	37.8 ± 0.9	10.3 ± 2.2	158.0 ± 198.4	33.4 ± 3.4	6.6 ± 10.1
21	222.2 ± 11.3	57.1 ± 0.5	23.3 ± 1.4	532.1 ± 54.5	54.11 ± 0.7	14.6 ± 1.7
22	238.6 ± 9.7	62.7 ± 0.5	26.7 ± 1.3	631.2 ± 39.5	61.8 ± 0.7	25.4 ± 1.8
23	222.9 ± 7.9	63.7 ± 0.5	31.3 ± 1.3	683.7 ± 41.1	59.1 ± 0.7	25.8 ± 1.8
24	120.1 ± 10.1	104.0 ± 0.8	20.3 ± 1.9	254.2 ± 22.2	107.1 ± 0.8	20.4 ± 2.0
25	71.5 ± 5.4	38.2 ± 2.0	54.5 ± 4.7	320.5 ± 17.1	32.0 ± 1.1	43.1 ± 2.6
26	70.7 ± 9.7	48.4 ± 1.8	26.6 ± 4.2	137.9 ± 21.0	42.5 ± 1.9	25.6 ± 4.5
27	77.6 ± 8.4	61.9 ± 1.5	28.2 ± 3.5	265.2 ± 25.1	60.4 ± 1.1	24.1 ± 2.6
	...	...	...	14.86 ± 11.2	−13.9 ± 1.9	50.8 ± 4.6
28	68.9 ± 5.9	34.1 ± 0.8	20.2 ± 2.0	190.6 ± 22.9	35.9 ± 1.4	24.9 ± 3.4
	49.6 ± 8.4	87.5 ± 2.0	24.3 ± 4.8	111.6 ± 17.9	87.8 ± 1.7	22.4 ± 4.2
29	128.1 ± 5.9	82.8 ± 0.5	23.5 ± 1.2	307.1 ± 37.6	84.8 ± 1.3	21.5 ± 3.2
30	36.3 ± 6.4	−11.5 ± 4.4	51.5 ± 10.5	...	...	...
31	199.1 ± 6.4	−11.7 ± 0.6	37.3 ± 1.4	613.3 ± 25.1	−10.8 ± 0.6	34.4 ± 1.6
32	228.7 ± 9.5	−22.6 ± 1.0	53.5 ± 2.5	730.6 ± 26.4	−19.5 ± 0.9	54.4 ± 2.2
33	250.5 ± 8.7	−22.4 ± 0.9	56.4 ± 2.2	831.9 ± 21.8	−23.1 ± 0.7	58.5 ± 1.7
34	211.8 ± 9.2	−23.3 ± 1.1	52.3 ± 2.6	807.3 ± 21.8	−22.3 ± 0.6	50.8 ± 1.5
35	152.9 ± 10.6	−104.6 ± 1.1	34.6 ± 2.7	620.2 ± 15.7	−103.4 ± 0.4	37.5 ± 1.1
36	64.5 ± 11.9	−81.7 ± 1.1	12.5 ± 2.6	183.6 ± 27.2	−79.3 ± 1.4	20.1 ± 3.4
	...	...	...	116.5 ± 34.1	−4.6 ± 3.1	2.20 ± 7.5
37	107.4 ± 11.6	−19.4 ± 1.3	24.9 ± 3.2	444.4 ± 42.9	−17.2 ± 1.1	24.4 ± 2.7

Download table as:  ASCIITypeset image

The SiO clumps are interpreted to be arising from protostellar outflow sources. Using a typical molecular density ∼10⁶ cm⁻³ and a radius of ∼2'', the total mass that is swept up a molecular clump of SiO emission by protostellar outflow is estimated to be ∼68 ${M}_{\odot }$. The kinetic luminosity ${L}_{\mathrm{kinetic}}=0.5\;{\mathrm{Mv}}^{2}$ of the swept-up material over the dynamical age ∼1300 years is also calculated to be ∼500 ${L}_{\odot }$. Using the estimated SiO and molecular hydrogen column densities, the abundance of SiO in each SiO clump is estimated to be 2.5 × 10⁻¹⁰. The column density of hydrogen itself is estimated by multiplying the gas density and the size of a clump. However, SiO shocks are generally associated with a thin layer where the protostellar jet interacts with a cloud. So, we overestimated the thickness of the SiO layer by a factor of 10 to 100 and therefore underestimated the SiO abundance. Thus, the SiO abundance could range between 2.5 × 10⁻⁹ and 2.5 × 10⁻⁸.

3.3.3. Water Maser, HCN, and SiO Lines

To investigate whether water masers (see Figure 1) are tracking star formation activity, we compare the distribution of SiO (5-4) emission with the spectra of several water masers toward the molecular ring. Figure 8 shows a grayscale image of SiO (5-4) line emission from the molecular ring and the insets show six SiO (5-4) spectra from regions where water masers have been detected. Water masers 1, 11, and 12 lie in the vicinity of HCN(1-0) clumps F and G, whereas water masers 4 and 5 are concentrated near HCN (1-0) clump O. These SiO features associated with clumps F, G, and O to the E and SW of the molecular ring are resolved spatially in the N–S direction. The SiO (5-4) spectra 21 to 24 from the HCN (1-0) clumps F and G reveal velocities peaking between 30 and 60 km s⁻¹. These velocities are generally consistent with those of water masers 1 and 12. The velocity profile of SiO (5-4) from clump V shows two blue and redshifted velocity components whereas HCN (1-0) spectra show a broad asymmetric redshifted wing. SiO (5-4) line emission from Clump O shows a velocity peak at ∼−8 km s⁻¹ and extends to negative velocity of −100 km s⁻¹. The broad line width of this clump includes the LSR velocities of water masers 4 and 5. The spectra of SiO (5-4) line emission from clump O exhibit an asymmetric blueshifted wing.

Zoom In Zoom Out Reset image size

Previous H₂ observations of the molecular ring (Gatley et al. 1986; Yusef-Zadeh et al. 2001) have identified shocked H₂ features associated with clumps V and O. The correspondence between SiO, HCN(1-0), and H₂ molecular line emission suggests that H₂ emission from the CMR is shock excited. The velocity profiles of SiO emission are similar to those of HCN, thus suggesting that SiO (5-4) and HCN (1-0) line emission arise from the same region. In addition, velocities of water masers 1, 12, 4, and 5 are within the range of SiO emission from clumps F, G, and O, thus suggesting that these water masers may be associated with the molecular gas in the ring.

Figure 9(a) shows contours of HCN (1-0) line emission at the peak velocity of clump V superimposed on a grayscale continuum image of the mini-spiral H ii feature at 5 GHz (Yusef-Zadeh & Wardle 1993). A radio continuum ionized feature, which we refer to as the EW Ridge, extends away from clump V at roughly constant declination. This ridge of ionized gas branches into two linear features, each with a horizontal and vertical extent of ∼20'' (0.8 pc) and 1'' (0.04 pc), respectively, crossing each other with PAs of 81° and 105° at the peak of clump V peaks. The kinematics of this feature do not follow the kinematics of the molecular ring. Morphologically, the linear features wiggle to the east as they merge with the N arm of the mini-spiral. Figure 9(b) shows the position–velocity (P–V) contour plot of Ne ii line emission along the brighter branch along the length of the EW ionized ridge at a PA of −93° at almost constant declination. This plot illustrates the gradual change in the velocity of the linear feature from ∼55 to −55 km s⁻¹ with a velocity gradient of ∼470 km s⁻¹ pc⁻¹. The red and blueshifted velocity components are associated with the eastern and western halves of the crossed linear features, respectively. Given the high-velocity gradient of Ne ii line emission, the ionized gas associated with the EW ridge is kinematically disturbed and is likely to be unbound to the Galactic center gravitational potential. The blue- and redshifted radial velocities along the linear features are reminiscent of bipolar outflows.

Zoom In Zoom Out Reset image size

We suggest that the EW feature is a result of jet activity from two protostars in clump V. A collisionally excited methanol maser is associated with clump V (YBRW08), at a velocity of ∼40 km s⁻¹, superimposed on broad thermal line emission extending to velocities of ∼80 km s⁻¹. There is no evidence of water masers arising from clump V. The dynamical age of the outflow is estimated to be ∼1 × 10⁴ years using the size of the ionized ridge ∼20'' (0.78 pc) and the difference in the blue and redshifted velocities ∼80 km s⁻¹. The intensity of the emission from the EW Ridge is ∼0.5 mJy beam⁻¹ at 22.4 GHz with a resolution of 024 × 017. The background subtracted flux density of the ionized EW Ridge is estimated to be ∼0.5 ± 0.2 Jy at 22.4 GHz. The ionized gas is estimated to have an emission measure EM ∼6 × 10⁶ pc cm⁻⁶ corresponding to an ionized gas density of 8.7 × 10³/f^0.5 cm⁻³ where f is the volume filling factor. The electron temperature is assumed to be ${T}_{{\text{}}e}\sim 8000\;{\rm{K}}$ and a path-length L = 0.08 pc, which is equivalent to the jet diameter ∼2''. The total mass of the EW Ridge and the mass-loss rate of the outflowing ionized material are estimated to be $\sim 1.7\times {f}^{0.5}$ ${M}_{\odot }$ and $\sim 1.8\times {f}^{0.5}\times {10}^{-4}$ ${M}_{\odot }$ yr⁻¹, respectively. Assuming 10% lumpiness in the density of ionized gas, f = 0.1, the total mass and the mass-loss rate are ∼0.5 ${M}_{\odot }$ and $\sim 5\times {10}^{-5}$ ${M}_{\odot }$ yr⁻¹, respectively. These values are generally consistent with those found in star-forming regions.

Star formation activity within the inner 2 pc of Sgr A* requires a reservoir of molecular gas surrounding the sites of young star-forming activity. The CMR provides the largest concentration of molecular gas near Sgr A*. Recent observations have also shown additional molecular material is distributed within the CMR, contrary to the widely accepted idea that a cavity of mini-spiral shaped ionized gas fills inside the molecular ring (Jackson et al. 1993; Monterrey-Castño et al. 2010; Martin et al. 2012; Yusef-Zadeh et al. 2013). Here we provide a new tracer of extended neutral gas by identifying radio dark clouds within the molecular ring. The presence of radio dark clouds is consistent with a reservoir of neutral gas that feeds star formation activity close to Sgr A*.

Recent radio images of the Galactic center have revealed a large number of radio dark clouds (Yusef-Zadeh 2012). These features coincide with cool neutral clouds embedded in a hot ionized medium, tracing volumes where there is a deficiency of free–free radio continuum emission. Figures 10(a) and (b) show the distribution of ionized gas surrounding Sgr A* at 7 mm (Yusef-Zadeh et al. 2014). We note three dark features, two near the N and E arms of the mini-spiral, and one near the IRS 13 complex. The darkest features lie to the east of IRS 13E and SE of IRS 13N, identified by a red circle with a diameter ∼1''. We suggest that these dark features are associated with radio dark clouds. Interferometric errors due to incomplete sampling of UV plane also produce dark features. However, the regions labeled in Figure 10 have been detected in radio images at multiple wavelengths based on different observations, suggesting that they are real features. Future high-resolution molecular line observations should confirm our interpretation of radio dark clouds. The mean flux density averaged over the red circle near IRS 13 is fainter by 0.1–0.2 mJy than the region immediately outside the circled region. The difference between the flux density of the ambient gas toward and away from the cloud suggests that the depth of the ionized layer is 5–10 times bigger than the dimension of the radio dark cloud. This radio dark cloud may be tracing the molecular material associated with ongoing star formation near IRS 13N. Similarly, the radio dark cloud near the N and E arms indicates that the ionized gas in the mini-spiral feature is well mixed with layers of molecular gas, as traced by dark radio features. In this case, the length of the ionized layer in the N and E arms is two to three times larger than the size of the clouds ∼8'' (∼0.3 pc) associated with the N and E arms.

Zoom In Zoom Out Reset image size