A SEARCH FOR SUBMILLIMETER H₂O MASERS IN ACTIVE GALAXIES: THE DETECTION OF 321 GHz H₂O MASER EMISSION IN NGC 4945

Maser emission at the H₂O ${6}_{16}\mbox{--}{5}_{23}$ transition (rest frequency ${\nu }_{\mathrm{rest}}$ = 22.23508 GHz) occurs over a wide range of physical conditions with kinetic temperatures of T_k = 200–2000 K, hydrogen densities of n(H₂) = 10⁸–10¹⁰ cm⁻³, and H₂O densities of n(H₂O) = 10³–10⁵ cm⁻³ (e.g., Elitzur et al. 1989; Neufeld et al. 1995; Humphreys et al. 2005). Extragalactic 22 GHz H₂O masers provide important information on the geometry and kinematics of the central few parsecs of active galactic nuclei (AGNs) in terms of dense molecular gas. Submillimeter H₂O maser transitions at the 321.226 GHz (${10}_{29}\mbox{--}{9}_{36}$) and 325.153 GHz (${5}_{15}\mbox{--}{4}_{22}$) transitions are strongly inverted under more restricted physical conditions (e.g., ${T}_{k}\gt 1000$ K; Deguchi 1977; Neufeld & Melnick 1990, 1991). The first detection of a 321 GHz (${10}_{29}\mbox{--}{9}_{36}$) H₂O maser in a galactic star-forming region was achieved by Menten et al. (1990a). Subsequent detections of (sub)millimeter H₂O masers at 183.308, 321.226, and 325.153 GHz in galactic sources (Menten et al. 1990a; Menten & Melnick 1991; Humphreys 2007) have inspired similar searches toward active galaxies.

The studies of extragalactic submillimeter H₂O masers were pioneered by Humphreys et al. (2005), who discovered extragalactic submillimeter H₂O masers at the 183 GHz (${3}_{13}\mbox{--}{2}_{20}$) and 439 GHz (${6}_{43}\mbox{--}{5}_{50}$) transitions toward the nuclear region of the type 2 Seyfert/LINER galaxy NGC 3079 using the Submillimeter Array (SMA) and James Clerk Maxwell Telescope. The detection of 183 GHz H₂O emission from the merging galaxy Arp 220, which hosts two nuclei but no active nucleus in its center, has also been reported (Cernicharo et al. 2006). Hagiwara et al. (2013) reported the first extragalactic detection of 321 GHz H₂O maser emission toward the center of the Circinus galaxy, which has a type 2 Seyfert nucleus, using the Atacama Large Millimeter/submillimeter Array (ALMA).

These studies suggest that extragalactic submillimeter H₂O masers may be common in AGNs and could be associated with AGN activity or trace a circumnuclear disk around the AGN like the nuclear H₂O megamasers at 22 GHz (e.g., Miyoshi et al. 1995), although the 183 GHz H₂O emission in Arp 220 arises from a star-forming site in the galaxy rather than AGN activity (Cernicharo et al. 2006). Following our previous study of 321 GHz maser emission in the Circinus galaxy, we present here the results of a search for 321 and 325 GHz H₂O masers toward several AGNs, conducted with ALMA. Throughout this article, cosmological parameters of H₀ = 73 km s⁻¹ Mpc⁻¹, ${{\rm{\Omega }}}_{{\rm{\Lambda }}}=0.73$, and ${{\rm{\Omega }}}_{M}=0.27$ are adopted. The optical velocity definition is adopted throughout this article, and the velocities are calculated with respect to the local standard of rest (LSR).

In this program, we concentrated our search for extragalactic submillimeter H₂O masers at the 321.226 GHz transition from the known H₂O nuclear masers with the strongest maser flux densities at 22 GHz (Dos Santos & Lépine 1979; Claussen & Lo 1986; Greenhill & Gwinn 1997; Braatz et al. 2003) and additionally searched for the 325.153 GHz transition from the megamaser galaxy NGC 5793 (Hagiwara et al. 1997). It is generally understood that 321 and 325 GHz H₂O masers exist in circumstellar envelopes in galactic star-forming regions (e.g., Menten et al. 1990a; Menten & Melnick 1991), while the 325 GHz H₂O emission is known to be strongly attenuated by water vapor absorption in the atmosphere. However, the attenuation of the 325 GHz H₂O line is less significant for distant galaxies due to the redshift of the H₂O line velocity, and so we included NGC 5793, at distance of 51 Mpc, in our list (Table 1). All the galaxies in our list contain 22 GHz nuclear masers which are subcategorized as disk masers, having characteristic spectra implying the presence of a disk around a nucleus (e.g., Greenhill et al. 2009). The presence of a nuclear edge-on disk in the nuclear region makes the detection of submillimeter H₂O masers more likely (e.g., Humphreys et al. 2005). However, it is also possible that submillimeter maser emission may arise from extragalactic star-forming sites like the 22 GHz H₂O masers in star-forming galaxies (e.g., Greenhill et al. 1993; Hagiwara 2007).

Table 1.  Summary of a Search for Extragalactic Submillimeter H₂O Emission

Source	R.A.a	decl.a	Db	Datec	$\nu$ d	N${}_{{\rm{A}}}$ e	t${}_{{\rm{on}}}$ f	${\theta }_{b}$ g (PAh)	$1\sigma$ i
	(J2000)	(J2000)	(Mpc)		(GHz)		(min)	(arcsec2, deg)	(mJy)
NGC 1068	02h42m40770	−00°00'4784	12.5	Jun 2–6	319.07	21	39	0.61 × 0.44 (42)	4–6
NGC 1386	03h36m46237	−35°59'5739	10.6	Jun 5	319.36	29	11	0.67 × 0.54 (−27)	10–15
NGC 4945	13h05m27279	−49°28'0444	11.1	Jun 3	319.68	18	14	0.54 × 0.51 (11)	8–11
Circinus	14h13m09906	−65°20'2046	4.2	Jun 3	319.82	18	18	0.66 × 0.51 (−17)	9–11
NGC 5793	14h59m24807	−16°41'3655	51.1	Jun 3	316.59	21	6	0.56 × 0.47 (47)	11–13
NGC 5793	14h59m24807	−16°41'3655	51.1	Jun 3	320.47	21	19	0.66 × 0.46 (−89)	8–11

Notes.

^aInterferometry phase center positions in R.A. and decl. ^bLuminosity distances, adopted from NED. ^cDate of observations in 2012. ^dObserving frequencies at channel 0 in each spectral window. ^eNumber of the 12 m antenna used in the ALMA observations. ^fOn-source time for the target source. ^gSynthesized beam size. ^hThe beam position angle. ⁱThe rms noise values in a 488.3 kHz spectral resolution in the clean image, depending on channel.

Download table as:  ASCIITypeset image

We conducted spectral-line observations at the 321 and 325 GHz transitions toward five galaxies in ALMA Band 7 during the Cycle 0 time between 2012 June 2 and 6 under the experiment code #2011.0.00121.S (PI: Hagiwara). Hagiwara et al. (2013) have previously reported the detection of 321 GHz H₂O maser emission in the nearby Circinus galaxy arising from this set of observations, and preliminary results from this fuller study were presented by Hagiwara et al. (2015). A summary of the observations is given in Table 1, including the target galaxies, observation dates, and the number of antennas employed in each observation. We set the detection threshold at the 5σ level, with the observed 1σ noise level of ∼20 mJy resulting in a flux-density limit of about 100 mJy beam⁻¹ for all sources. This was considered to be sufficient to search for new submillimeter H₂O masers from the known 22 GHz megamaser galaxies having a disk or disk-like structure. The durations of the observations ranged from 20 to 50 minutes (approximately 5–20 minutes on-source), depending on the target source. In all our observations, a single dual-polarization spectral window (1.875 GHz bandwidth) in the Frequency Division Mode was employed: the single 1.875 GHz bandwidth was divided into 3840 spectral channels, yielding spectral resolutions of 488 kHz or ∼0.45 km ${{\rm{s}}}^{-1}\$ at the observed frequency of 319.4–321.5 GHz. The resultant total velocity coverage was ∼1760 km s⁻¹, centered near each galaxy's systemic velocity. We conducted phase-referencing observations using a bright phase-reference source within ∼10° of the target sources. Data calibration was performed using the Common Astronomy Software Applications (CASA). Amplitude calibration was performed using observations of Titan or Uranus, and the bandpass calibration was made using the bright quasar, either 3C 279 or 3C 454.3, that was nearest to the target source. Flux-density errors of 10% are adopted, as stated in the capabilities of Cycle 0 observations for Band 7. The image analysis was performed using both CASA and the Astronomical Image Processing Software. After phase and amplitude calibration, the 321/325 GHz continuum emission was subtracted from the spectral line visibilities using line-free channels prior to the imaging and CLEAN deconvolution. The line emission from each galaxy was separated out from the continuum emission. The synthesized beam sizes (natural weighting) and the beam position angles used in the CLEAN process and the resultant rms noise levels in the spectral line images are listed in Table 1.

We searched for maser emission in each AGN within a field of view of ∼18'', centered on the phase-tracking center of each source and within the observed velocity range. Figure 1 shows spectra at 321 GHz from the four AGNs, NGC 1068, NGC 1386, NGC 4945, and NGC 5793, and at 325 GHz from NGC 5793. The only emission exceeding a signal-to-noise ratio (S/N) of 5 is seen in NGC 4945, as reported by Hagiwara et al. (2015). The detection of H₂O emission toward NGC 4945 is reminiscent of that toward the Circinus galaxy reported in our earlier study (Hagiwara et al. 2013). Given the narrow line widths of the H₂O emission and the fact that it is unresolved at the angular resolution of our observations, it is most likely that the detected H₂O emission has a maser origin, as in the case of Circinus. Figure 2 displays the spectrum of the H₂O maser emission and the 321 GHz continuum emission in NGC 4945. Several 321 GHz H₂O maser features were detected in NGC 4945, with peak flux densities of ∼45 mJy beam⁻¹ at V${}_{\mathrm{LSR}}$ = 714.4 km ${{\rm{s}}}^{-1}$ and ∼50 mJy beam⁻¹ at V${}_{\mathrm{LSR}}$ = 718.9 km s⁻¹: all these features are detected with an S/N > 5. In Figure 1, the detected H₂O emission of NGC 4945 comprises distinct features that are redshifted from the systemic velocity of the galaxy (V${}_{\mathrm{LSR}}$ = 556 km s⁻¹). The feature at V${}_{\mathrm{LSR}}$ = 714.4 km ${{\rm{s}}}^{-1}\$, which shows the strongest peak, is redshifted by 162.9 km ${{\rm{s}}}^{-1}\$ from the systemic velocity. Most of the other prominent features lie in the redshifted velocity range of V${}_{\mathrm{LSR}}$ = 711–722 km ${{\rm{s}}}^{-1}\$, with the exception of a narrow feature at V${}_{\mathrm{LSR}}$ = 679.9 km ${{\rm{s}}}^{-1}\$ with a peak flux density of ∼40 mJy beam⁻¹. The total integrated intensity estimated from these features lying at V${}_{\mathrm{LSR}}$ = 679–680 km ${{\rm{s}}}^{-1}\$ and V${}_{\mathrm{LSR}}$ = 712–722 km ${{\rm{s}}}^{-1}\$ is ∼2.6 Jy km s⁻¹, which corresponds to ∼7.4 L${}_{\odot }$. We note a highly redshifted feature at the ∼3σ level centered at V${}_{\mathrm{LSR}}$ = 1138.1 km ${{\rm{s}}}^{-1}\$ with a peak strength of ∼35 mJy beam⁻¹(Figures 1, 2). The corresponding fringe phases do not clearly show a consistent trend. Further, more sensitive observations are required to confirm this feature.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

In Figure 3, we compare the 321 GHz H₂O spectrum with a 22 GHz spectrum obtained on 2014 April 7 using the Deep Space Network 70 m antenna at Tidbinbilla. This observation was conducted at a spectral resolution of 31.25 kHz or 0.42 km ${{\rm{s}}}^{-1}\$ and with an rms of ∼0.3 Jy. We conservatively estimate that the uncertainty in amplitude is up to 50%. The known strong variability of the 22 GHz maser emission in this galaxy, together with the ∼1.8 yr between the two observations, makes it difficult to precisely compare the velocities of each feature between the 22 and 321 GHz lines. In our 321 GHz observation, no feature was detected at or near the systemic velocity, while the 22 GHz spectrum in Figure 3 shows a systemic feature, which is consistent with earlier single-dish measurements (e.g., Greenhill et al. 1997). It should be noted that an asymmetry between the blueshifted and redshifted emission is commonly observed in H₂O megamasers, such as NGC 4258: redshifted features are more numerous and are significantly more intense than the blueshifted features (e.g., Maoz & McKee 1998).

Zoom In Zoom Out Reset image size

A comparison of the 321 GHz H₂O spectrum with the 22 GHz spectrum in Figure 3 and similar spectra in the literature (Nakai et al. 1995; Braatz et al. 2003) indicates that the detected 321 GHz emission in the galaxy is similar to the features appearing in the 22 GHz maser spectra spanning V${}_{\mathrm{LSR}}$ = 643–714 km s⁻¹; however, there is no direct correspondence between the velocities of the features in the two bands. In addition, in the published 22 GHz maser spectrum of Greenhill et al. (1997), the most redshifted velocity feature at V${}_{\mathrm{LSR}}$ = 774.4 km ${{\rm{s}}}^{-1}\$ was tentatively detected, whereas there is no corresponding feature in either our 321 GHz or 22 GHz spectra. (Velocity values are converted from optical heliocentric to the LSR definition, using the relation V${}_{\mathrm{LSR}}$ = V_hel − 4.6 km ${{\rm{s}}}^{-1}\$ in Greenhill et al. 1997.) Based on the ALMA data, the Gaussian-fitted phase-referenced position of the V${}_{\mathrm{LSR}}$ = 718.0 km ${{\rm{s}}}^{-1}\$ peak emission is α(2000) = 13^h05^m2748, δ(2000) = −49°28'0550. Positional uncertainties of 003 are estimated. The position of the marginally significant redshifted emission at V${}_{\mathrm{LSR}}$ = 1138.6 km ${{\rm{s}}}^{-1}\$ coincides with this peak within the uncertainties. According to the ALMA Cycle 0 capabilities, the positional errors are ∼10% of the synthesized beam, corresponding to 054/10 ∼0054 in our observation, and this can be taken as the most dominant source of error. The relative positions of other redshifted features coincide within uncertainties, and the features remain unresolved at the angular resolution of ∼054, or ∼30 pc. The position of the 22 GHz H₂O maser (α(2000) = 13^h05^m2748 ± 0.02, δ(2000) = −49°28'054 ± 0.1; Greenhill et al. 1997) coincides with that of the 321 GHz maser within uncertainties.

A 321 GHz submillimeter continuum image is obtained toward the nucleus of the galaxy, with a Gaussian-fitted peak position of α(2000) = 13^h05^m27.47, δ(2000) = $-49^\circ 28^{\prime} 05\buildrel{\prime\prime}\over{.} 42$ (with uncertainties of ∼0015). The continuum peak flux density at this resolution is 136.5 mJy beam⁻¹, which is detected with an S/N of ∼20 (Figure 2) and an rms noise of about 7 mJy beam⁻¹, with a total continuum flux density of 543.3 mJy beam⁻¹. Thus, the approximate relative positional errors between the H₂O emission and the continuum peak, represented by the synthesized beam size divided by 2 × S/N (Hagiwara et al. 2001), are ∼0055 or 3.0 pc, within which the maser resides in the submillimeter continuum nucleus. Therefore, we conclude that the locations of the 321 GHz H₂O maser spots coincide with that of the submillimeter continuum peak within uncertainties of ∼3.0 pc.

5.1. The 321 GHz Continuum and New Maser Emission in NGC 4945

5.2. Submillimeter H₂O Masers in AGNs

5.3. High-velocity Dense Gas in NGC 4945

This research was supported by the Japan Society for the Promotion of Science (JSPS) Grant-in-Aid for Scientific Research (B) (Grant Number: JP15H03644). This article makes use of the following ALMA data: #2011.0.00121.S. ALMA is a partnership of ESO (representing its member states), NSF (USA), and NINS (Japan), together with NRC (Canada) and NSC and ASIAA (Taiwan), in cooperation with the Republic of Chile. The Joint ALMA Observatory is operated by ESO, AUI/NRAO, and NAOJ. This research has made use of the NASA/IPAC Extragalactic Database (NED), which is operated by the Jet Propulsion Laboratory, California Institute of Technology, under contract with the National Aeronautics and Space Administration. The Tidbinbilla 70 m telescope is part of the NASA Deep Space Network and is operated by CSIRO.

Facilities: ALMA - Atacama Large Millimeter Array.