SIGNATURE OF AN INTERMEDIATE-MASS BLACK HOLE IN THE CENTRAL MOLECULAR ZONE OF OUR GALAXY

Most galaxies, including the Milky Way, are thought to have black holes (BHs) with masses greater than a million solar masses (M_⊙) at their centers. However, the origins of such supermassive BHs (SMBHs) remain unknown (e.g., Djorgovski et al. 2008). One possible scenario is that a "seed" BH with tens or hundreds of solar masses at the center of a galaxy grows by accretion of matter (e.g., Volonteri & Rees 2005). Another scenario is that BHs with masses of ∼10³ M_⊙, which are formed by runaway coalescence of stars in young compact star clusters (Portegies Zwart et al. 1999), merge at the center of a galaxy to form an SMBH (Ebisuzaki et al. 2001). The latter theory is slightly more persuasive, as it naturally explains the tight correlation between the mass of a given central SMBH and that of the associated galactic bulge. To confirm the merging scenario, the ability to unambiguously detect intermediate-mass BHs (IMBHs; M = 10^2–5 M_⊙) is essential. Many candidates for IMBHs have been proposed to date on the basis of their ultraluminous nature (Fabbiano 2006; Roberts 2007), low-temperature blackbody spectral components (Miller et al. 2003), and quasi-periodic oscillations (Casella et al. 2008). One promising IMBH candidate is ESO 243-49 HLX-1, which is the brightest ultra-luminous X-ray source (ULX) outside galactic nuclei (Godet et al. 2009). However, the nature of ULXs is uncertain, and significant evidence exists that they are not IMBHs (e.g., Ebisawa et al. 2003; Liu & Mirabel 2005; Okajima et al. 2006). IMBHs have been inferred in several globular clusters from velocity dispersion profiles (Gebhardt et al. 2002; Gerssen et al. 2002; Noyola et al. 2008), but they are often not confirmed by other teams (e.g., Strader et al. 2012). None of these IMBH candidates are widely accepted as definitive, thus their existence is a long-standing controversy. In this Letter, an alternative approach to seeking IMBHs based on analysis of the kinematical structures of certain compact clouds is proposed.

The nucleus of the Milky Way Galaxy (Sgr A*) also harbors a 4 million M_⊙ BH (e.g., Ghez et al. 2008; Gillessen et al. 2009). A large expanse of dense interstellar gas exists within several hundred parsecs of the nucleus, which is known as the central molecular zone (CMZ; Morris & Serabyn 1996). While investigating CO J = 3–2 survey data of the Galactic CMZ obtained with the Atacama Submillimeter Telescope Experiment (ASTE) 10 m telescope (Oka et al. 2012), we noticed a peculiar molecular cloud at galactic longitude −040 and galactic latitude −022, with local standard of rest (LSR) velocities ranging from −120 to −20 km s⁻¹. This CO–0.40–0.22 is a compact cloud (<5 pc at 8.3 kpc distance) with an extremely broad velocity width (∼100 km s⁻¹) and a very high CO J = 3–2/J = 1–0 intensity ratio (≥1.5; Figure 1). The high ratio implies that the cloud consists of dense, warm, and moderate opacity gas (Oka et al. 2007, 2012). It belongs to a peculiar category of molecular clouds, namely, high-velocity compact clouds (HVCCs)³ , which were originally identified in the CO J = 1–0 emission survey data (Oka et al. 1998). The position of CO–0.40–0.22 is approximately 02 Galactic southeast of the massive star-forming region Sgr C, being displaced by ∼60 pc in projected distance from the Galactic nucleus. Its compact appearance and broad velocity width are remarkable in the HCN J = 4–3 emission image. It is located at the rim of an expanding shell of dense molecular gas (shell 2; Tanaka et al. 2014). Although an Hα emission region 03 × 03 in size, RCW137 (Rodgers et al. 1960), overlaps CO–0.40–0.22 in the plane of the sky, it has a positive LSR velocity and is therefore physically unrelated to CO–0.40–0.22.

Zoom In Zoom Out Reset image size

We performed mapping observations of CO–0.40–0.22 in 21 molecular lines in the 3 mm band using the Nobeyama Radio Observatory (NRO) 45 m radio telescope. Table 1 summarizes the frequencies, molecules, transitions, and parameters calculated for each line. These lines were selected to trace CO–0.40–0.22 on the basis of the result of a 3 mm band line survey toward this cloud with the Mopra 22 m telescope (Oka et al. 2014). The observations were made on 2014 January 25–February 3 and March 25–28.

Table 1.  Observed Line Parameters

Freq.	Molecule	Transition	TMBa	Sb	σV	e	P.A.
(GHz)			(K)	(pc)	(km s−1)		(°)
81.881	HC3N	J = 9–8	2.03 ± 0.10	1.17	21.6	0.66	42.9
85.339	c-C3H2	212–101	0.30 ± 0.06	1.19	29.4	0.33	−42.5
86.340	H13CN	J = 1–0	1.86 ± 0.06	1.33	23.8	0.56	44.2
86.754	H13CO+	J = 1–0	0.32 ± 0.07	1.50	24.8	0.65	40.1
86.847	SiO	J = 2–1	2.26 ± 0.08	1.19	19.6	0.66	42.7
88.632	HCN	J = 1–0	7.58 ± 0.09	1.58	26.0	0.41	44.0
89.189	HCO+	J = 1–0	2.48 ± 0.25	1.49	20.1	0.61	46.0
90.664	HNC	J = 1–0	2.19 ± 0.09	1.58	26.3	0.48	44.4
90.979	HC3N	J = 10–9	2.17 ± 0.09	1.10	20.8	0.67	43.6
93.174	N2H+	J = 1–0	2.53 ± 0.09	1.31	23.7	0.63	45.5
94.405	13CH3OH	2−1–1−1E	0.28 ± 0.07	1.45	27.3	0.44	46.2
95.169	CH3OH	800–710A	1.56 ± 0.09	0.73	15.0	0.59	44.3
95.914	CH3OH	210–110A	0.45 ± 0.11	1.56	28.2	0.23	−47.1
96.413	C34S	J = 2–1	0.50 ± 0.07	1.23	24.2	0.26	−40.6
96.741	CH3OH	200–100A	5.79 ± 0.09	1.36	25.4	0.57	44.7
97.981	CS	J = 2–1	5.42 ± 0.06	1.54	28.7	0.40	44.1
99.300	SO	32–21	1.83 ± 0.08	1.31	23.9	0.56	43.8
100.076	HC3N	J = 11–10	2.20 ± 0.12	0.94	19.6	0.71	43.3
101.478	H2CS	313–212	0.30 ± 0.11	⋯	⋯	⋯	⋯
102.064	NH2CHO	514–414	0.04 ± 0.14	⋯	⋯	⋯	⋯
103.040	H2CS	303–202	0.18 ± 0.12	⋯	⋯	⋯	⋯

Notes.

^aRefers to the T_MB at (l, b, V_LSR) = (−040, −022, −80 km s⁻¹). ^bThe size parameter is calculated as $S=D\;{\rm{tan}}\left((,\sqrt{{\sigma }_{l}{\sigma }_{b}},)\right)$ (Solomon et al. 1987), where D is the distance to the cloud (=8.3 kpc), and σ_l and σ_b are the dispersions along the Galactic longitude and latitude, respectively. Only the data with T_MB ≥ 3σ are used to calculate the dispersions (σ_V as well).

Download table as:  ASCIITypeset image

We used the TZ1 V/H receivers, which were operated in the two-sideband mode. The system noise temperatures ranged from 120 to 360 K at EL = 70° during the observations. We observed the 3' × 3' area around CO–0.40–0.22 in the on-the-fly mapping mode. All data were obtained by position switching between target scans and the clean reference position, (l, b) = (0°, −1°). The antenna temperature was calibrated by the standard chopper-wheel method. We used the SAM45 spectrometers in the 1 GHz bandwidth (244 kHz resolution) mode. The frequency resolution corresponds to a 0.8 km s⁻¹ velocity resolution at 86 GHz. The half-power beamwidth of the NRO 45 m telescope is ≃20'' at 86 GHz. Pointing errors were corrected every 2 h by observing the SiO maser source VX Sgr at 43 GHz. The pointing accuracy of the telescope was good to ≤3'' in both azimuth and elevation.

The obtained data were reduced using the NOSTAR reduction package. We subtracted the baselines of all spectra by fitting first- or third-order polynomial lines. We scaled the antenna temperature by multiplying it by 1/η_MB to obtain the main-beam temperature, T_MB. The main-beam efficiencies of the TZ1 V/H receivers at the line frequencies were calculated as ${\eta }_{{\rm{MB}}}(\nu )=a\;{\rm{exp}}\left((,-{(\nu /b)}^{2},)\right)$, where a and b were obtained by least-squares fitting to those measured at three frequencies (86, 110, and 115 GHz). All the data were resampled onto a 75 × 75 × 2 km s⁻¹ grid to obtain the final maps.

Eighteen of the twenty-one lines were detected from the center of CO–0.40–0.22 (Table 1). We calculated the size and velocity dispersion of the cloud using the data cube of the detected lines. We used the distance to the cloud D = 8.3 kpc. All the detected lines show that CO–0.40–0.22 has a compact appearance (S < 1.6 pc). The broad velocity width (σ_V > 20 km s⁻¹) is also common, except for the CH₃OH 8₀₀–7₁₀A line, which has a high upper-state energy (E_u/k ≃ 84 K). The HC₃N J = 11–10 line also shows a narrower velocity width and more compact appearance compared to the lower-J transitions of the same molecule. In Table 1, we also included the ellipticity (e) and the position angle (P.A.) with respect to the Galactic plane. Except for three low intensity lines (c-C₃H₂, CH₃OH, 2₁₀–1₁₀A, and C³⁴S), the P.A. of the cloud is concentrated around 45°, while the ellipticity ranges from 0.40 to 0.71.

We present the velocity-integrated map and a position–velocity map of the SiO J = 2–1 line, as it represents the spatial velocity structure of CO–0.40–0.22 (Figure 2). CO–0.40–0.22 has an oval shape with the major axis aligned with the shell 2. It consists of an intense component with a shallow velocity gradient and a less intense high-velocity wing. Shock probe lines, such as the SiO, SO, CH₃OH, and HCN J = 4–3 lines, commonly show this behavior. Neither an expanding feature nor a cavity appears in CO–0.40–0.22.

Zoom In Zoom Out Reset image size

A gas mass of M_gas = 10^3.6 M_⊙ is derived from the HCN J = 1–0 line intensity using the large velocity gradient model. Here, we assumed the kinetic temperatre to be T_k = 60 K which is derived for another well-studied HVCC CO 0.02–0.02 (Oka et al. 1999), and the [HCN]/[H₂] abundance ratio of 10^−7.3 which is derived for the Sgr B1 cloud (Tanaka et al. 2009; Oka et al. 2011). The density was chosen to be n(H₂) ≥ 10^6.5 cm⁻³ to reproduce the observed HCN J = 1–0 and J = 4–3 intensities at (l, b, V_LSR) = (−0.40, 0.22, −80 km s⁻¹), ${T}_{{\rm{MB}}}^{1-0}=7.6$ K, and ${T}_{{\rm{MB}}}^{4-3}=16.3$ K. A size parameter of 1.0 pc and velocity dispersion of 20 km s⁻¹ give a virial theorem mass of M_VT = 1 × 10⁶ M_⊙. This yields a very large virial parameter, M_VT/M_gas ∼ 300, indicating that the gas mass is definitely insufficient to bind the cloud by its self-gravity. The kinetic energy amounts to E_kin = 10^49.7 erg if the velocity dispersion is dominated by random motion. If it is expanding at V_exp = 40 km s⁻¹ (half of the velocity extent), the kinetic energy becomes E_kin = 10^49.8 erg. CO–0.40–0.22 is characterized by a rather featureless spatial velocity structure. This is in sharp contrast with CO 0.02–0.02 (Oka et al. 1999, 2008) or CO 1.27+0.01 (Oka et al. 2001; Tanaka et al. 2007), which are HVCCs containing expanding shells or emission cavities.

4.1. Origin of CO–0.40–0.22

4.2. Gravitational Kick Model

A promising candidate for an explanation of the formation scenario involves a "gravitational kick" to a small incoming cloud by a compact source. A point-like mass of 10⁵ M_⊙ can accelerate a cloud coming from infinity to ∼90 km s⁻¹ at a distance of 0.1 pc. This scenario can easily explain the compact appearance and very large velocity width of CO–0.40–0.22. However, we must then accept the presence of a massive compact (<0.1 pc) object. The association with the expanding shell supports this scenario because it can provide the massive object with a number of small clouds. For a semiquantitative comparison, we simulated the position–velocity behavior of an incoming cloud.

We placed a cloud of 200 test particles with a σ = 0.2 pc Gaussian centered at (ΔX, ΔY) = (10 pc, Y₀) offset from the massive object. The initial velocity was set to (v_X, v_Y) = (−10 km s⁻¹, 0 km s⁻¹). In this simulation, the free parameters are the central mass (M), initial Y value (Y₀), line of sight angle (ϕ), and elapsed time (t). The parameter ranges we searched are M = 10³–10⁶ M_⊙, Y₀ = (−3.0)–(−0.5) pc, ϕ = 0°–90°, and t = 0–10⁶ years. After a number of trials, we found that the parameter set M = 10⁵ M_⊙, y₀ = (−1.8)–(−1.0) pc, ϕ = 45°, and t ∼ 7 × 10⁵ years reproduces the position–velocity behavior of CO–0.40–0.22 well. Figure 3 shows the time evolution of two initially Gaussian clouds of σ = 0.2 pc with Y₀ = (−1.0) pc (cloud 1) and Y₀ = (−1.8) pc (cloud 2). The locus of the gravitational source corresponds to the center of CO–0.40–0.22, from which the positive-side highest velocity emission arises.

Zoom In Zoom Out Reset image size

What is this massive compact object? Supposing the gravitational kick scenario is valid, the object must be as large as 10⁵ M_⊙ with a radius that is significantly smaller than 0.1 pc (pericenter distance). These mass and size values correspond to an average mass density of ρ ≃ 2 × 10⁷ M_⊙ pc⁻³. This mass density is comparable to that of the core of M15, which is one of the most densely packed (core-collapsed) globular clusters in the Milky Way Galaxy (Djorgovski & King 1984). However, the total mass within 0.05 pc of the center of M15 is only 3400 M_⊙ (van den Bosch et al. 2006); thus, the generation of CO–0.40–0.22 by a globular cluster seems implausible. The lack of counterparts at other wavelengths is inconsistent with the massive stellar cluster interpretation, unless the cluster consists almost entirely of dark stellar remnants, such as neutron stars and BHs. Therefore, it is most likely that the massive compact object responsible for the formation of CO–0.40–0.22 is an invisible point-like mass of ∼10⁵ M_⊙, i.e., an IMBH.

The chance probability of a BH-molecular cloud encounter can be estimated by the gas volume filling factor. The volume filling factor of dense [n(H₂) ≥ 10⁴ cm⁻³] molecular gas was estimated to be >0.1 (Morris & Serabyn 1996). We expect the larger volume filling factor for the less dense (<10⁴ cm⁻³) molecular gas and the gas compression due to the encounter with a BH. Thus several IMBHs in the CMZ are sufficient to account for the detection of such an object.

4.3. Origin of the IMBH

This paper is based on observations conducted using the Nobeyama Radio Observatory (NRO) 45 m telescope and ASTE. The NRO is a branch of the National Astronomical Observatory of Japan (NAOJ), National Institutes of Natural Sciences. We are grateful to the NRO staff and all the members of the ASTE team for operation of the telescope. Observations with ASTE were conducted remotely from NRO using NTT's GEMnet2 and its partner research and education (R&E) networks, which are based on the AccessNova collaboration between the University of Chile, NTT Laboratories, and the NAOJ. We also thank F. Yusef-Zadeh for suggesting the gravitational kick scenario. T.O. acknowledges support from JSPS Grant-in-Aid for Scientific Research (C) No. 24540236.