A&A 383, 65-70 (2002)
DOI: 10.1051/0004-6361:20011687

A search for extragalactic H $\mathsf{_2}$O maser emission towards
IRAS galaxies

Detection of a maser from an infrared-luminous merger, NGC6240

Y. Hagiwara1 - P. J. Diamond 2 - M. Miyoshi 3


1 - Max-Planck-Institut für Radioastronomie (MPIfR) Auf dem Hügel 69, 53121 Bonn, Germany
2 - Jodrell Bank Observatory, University of Manchester, Macclesfield, Cheshire SK11 9DL, UK
3 - National Astronomy Observatory, Osawa, Mitaka, Japan

Received 31 July 2001 / Accepted 26 November 2001

Abstract
We report the result of an on-going survey for 22 GHz H2O maser emission towards infrared luminous galaxies. The observed galaxies were selected primarily from the IRAS bright galaxy sample. The survey has resulted in the detection of one new maser. The new maser was discovered towards the [U]LIRG/merger galaxy NGC6240, which contains a LINER nucleus. This is the first detection of an H2O$\;$maser towards this class of galaxy, they are traditionally associated with OH megamaser sources. The detected maser emission is highly redshifted ($\sim$260-300 km s-1) with respect to the adopted systemic velocity of the galaxy, and we identified no other significant emission at velocities $\mathrel{\mathchoice {\vcenter{\offinterlineskip\halign{\hfil $\displaystyle ...$\pm$500 km s$^{-1}\;$ relative to the systemic velocity. The presence of high-velocity maser emission implies the possible existence of a rotating maser disk formed in the merging process. The large maser luminosity ($\sim$40 $L_{\odot}$) suggests that an active galactic nucleus could be the energy source that gives rise to the water emission. Alternatively, the maser emission could be associated with the previously observed double radio source in the centre of the galaxy. Interferometric observations with high angular resolution will be able to clarify the origin of the new maser.

Key words: masers - galaxies: active - galaxies: individual (NGC 6240): radio lines - ISM: molecules


1 Introduction

The discovery of the highly Doppler-shifted maser emission at 22 GHz in the 616-523 transition of H2O$\;$in NGC4258 (Nakai et al. 1993), has resulted in a resurgence of single-dish surveys for extragalactic water masers. Recent maser surveys towards active galactic nuclei (AGN) have been motivated by the VLBI observations of a sub-parsec-scale H2O$\;$maser disk rotating around a central massive object in NGC4258 (Miyoshi et al. 1995). The detection and imaging of such H2O$\;$megamaser emission enables us to probe the kinematics and dynamical structures of parsec-scale circumnuclear disks (Moran et al. 1999). H2O$\;$megamasers studied to date can be categorised into two types; disk masers residing in a (sub-)parsec scale disk around an active nucleus, and non-disk masers which are seen significant distances from the nucleus. The latter can also be sub-categorised into two types: jet masers associated with radio jets e.g. NGC1068 (Gallimore et al. 1996), NGC1052 (Claussen et al. 1998), and Mkn348 (Falcke et al. 2000; Peck et al. 2001; Xanthopoulos & Richards 2001) or a nuclear outflowing component such as that observed in the Circinus galaxy (Greenhill 2000). Masers associated with jets appear to have different spectral characteristics than the disk masers, the spectra tend to be broad (a few hundred km s-1) and featureless, whereas the disk maserstend to consist of bright, narrow components clustered in velocity groups. H2O$\;$megamasers generally seem to trace the nuclear activity in the central parsecs of AGN. To deduce the general properties of such megamasers and those of the central parsecs of AGN, we need to increase the number of detections of water megamasers.
 

 
Table 1: Observed galaxies.
  Source   $\alpha$a   $\delta$a    $V_{\rm sys}$ a,b $\Delta V^{c}$ rms Epoch
  (1950) (1950) (km s$^{-1}\;$) (km s$^{-1}\;$) (mJy)  
UGC 05101 09 32 05 +61 34 37 11809 11580-12070 12 A
NGC 3310 10 35 40 +53 45 49 993 500-1520 21 A
ARP 299 11 25 38 +58 51 14 3111 2880-3370 18 C
NGC 3822f 11 39 36 +10 33 19 6120 5340-6610 16 A
NGC 4388 12 23 15 +12 56 20 2524 2290-2780 28 B
NGC 4438 12 25 14 +13 17 07 71 -160-340 32 B
NGC 4490 12 28 10 +41 55 08 565 330-820 14 A
NGC 4532 12 31 47 +06 44 39 2012 1780-2270 25 B
NGC 4631 12 39 42 +32 48 52 606 380-860 12 A
NGC 6052 16 03 01 +20 40 37 4716 4490-4990 14 A
NGC 6090 16 10 24 +52 35 04 8785 8560-9040 25 B
NGC 6240 16 50 28 +02 28 58 7339 6850-7870 4.5-7.0 B, C, D
NGC 6285 16 57 37 +59 01 47 5580 5190-6220 14 A
NGC 7674 23 25 24 +08 30 13 8671 8210-9230 28 A
UGC 12914 23 59 08 +23 12 58 4371 3890-4910 14 A


a From NED database.
b Heliocentric velocity.
c Observed velocity range.
d Observing epoch; A(23-24 April), B(9-10 May), C(13-14 June), and D(16-17 July) all in 2001.
e Condon et al. (1990).
f The source was not selected from IRAS galaxies.


At this time, 24 extragalactic water masers inside AGN have been discovered with single-dish surveys (e.g. Moran et al. 1999). The detection rates of new masers remain quite low, less than $\sim$7% (e.g., Braatz et al. 1996; Braatz et al. 1997; Greenhill et al. 1997). From the presently known sample of sources we have already deduced some properties of H2O$\;$megamasers. They prefer to lie in type 2 Seyfert AGN, suggesting that obscuring gas around an active nucleus coupled with long gain paths along an edge-on disk in the line of sight, plays a vital role in giving rise to strong maser emission. A number of H2O$\;$megamasers are known to contain a parsec-scale radio "core-jet'' structure which the maser can amplify in the gaseous environment. Accordingly it is reasonable to search for H2O$\;$megamasers towards type 2 Seyfert/LINER nuclei enshrouded by circumnuclear gas around a compact continuum nucleus.

Here we report the result of an on-going single-dish survey for H2O$\;$megamasers towards AGN in infrared (IR)-luminous galaxies, and the discovery of the new maser in the galaxy NGC6240.

2 Sample selection

The observed galaxies were primarily selected from far-infrared (FIR) luminous galaxies (IRAS galaxies) with an apparent FIR luminosity ( $L_{\rm FIR}) > 10^{11}~L_{\odot}$(Soifer et al. 1989). Condon et al. (1990) carried out 1.49 GHz imaging with the VLA at arcsecond resolution for these objects. Due to the now well-known correlation of $L_{\rm FIR}$ with radio continuum flux density for the IR-luminous galaxies, Condon et al.'s work has resulted in a significant number of detections of radio continuum sources towards such galaxies. The sample was selected from galaxies in which the ratio of 1.49 GHz continuum flux density to far-infrared (FIR) flux density ( $F_{\rm 21\,cm}/F_{\rm FIR}$) was >0.003, where $F_{\rm FIR} = F_{\rm 60\,\mu m} + F_{\rm 100\,\mu m}$. About 25 IR-luminous galaxies were selected in this way, of which 14 sources have been observed to date. Here we present the initial results of the survey. The sample observed so far is listed in Table 1.

3 Observations

The observations of H2O$\;$emission (rest frequency: 22.23508 GHz) were made with the MPIfR 100 m radio telescope at Effelsberg between April and July 2001, as part of an extragalactic H2O$\;$maser survey towards infrared-luminous galaxies. 15 galaxies have been observed to date (Table 1). We used a K-band HEMT receiver with two orthogonal linear polarizations. The system temperature was typically 70-110 K, depending on atmospheric conditions. The amplitude calibration was based on a flux density at 22 GHz for 3C286 (Ott et al. 1994), and flux densities were estimated from the measured antennas temperature, ${T_{\rm A}}^{*}$. The observations were made in position-switching mode. Typical observing time for each source was about 50-80 min, yielding rms sensitivities of $\sim$15-30 mJy. Pointing calibrations using nearby calibrator sources were carried out during the observations every $\sim$1-1.5 hours. The resultant pointing accuracy was <8 $^{\prime\prime}$ i.e. within 20% of the antenna beam size (40 $^{\prime\prime}$ [FWHM]). Uncertainties in the amplitude calibration are nominally 10-20%. The back-end digital autocorrelation system provided 4 autocorrelators for each of the 2 polarisation signals. Each autocorrelator produced a spectrum in a 40 MHz band (corresponding to a rest-frame velocity range of 540 km s-1) with 512 spectral channels providing a frequency resolution of 78 kHz or 1.1 km s-1. The center frequencies of the four autocorrelators were arranged to have 10 or 20 MHz overlap (depending on source), resulting in a total velocity coverage of $\pm$500 km s$^{-1}\;$relative to the galaxy's systemic velocity. After adding the two polarizations together, first or second order baselines were fitted to, and subtracted from, each of the spectra.

4 The survey results

The survey of 15 sources resulted in the detection of one new H2O$\;$megamaser. On 9 May 2001, we detected H2O$\;$emission towards the LINER galaxy NGC6240 (Fig. 1a), the detection was confirmed on 14 Jun. 2001 (Fig. 1b). We searched for water masers in the other 14 AGN but obtained no detections at a 3 $\sigma$ level of $\sim$40-90 mJy (Table 1). The detection of one new maser among the 15 AGN in the biased sample yields a detection rate of $\sim$6.7%, which is comparable to those of $\sim$7.0% in Braatz et al. (1997) and $\sim$3.4% in Greenhill et al. (1997). The H2O$\;$maser in the Seyfert 2 galaxy NGC5793 was originally selected from this sample, its discovery stimulated us to continue the programme (Hagiwara et al. 1997; Hagiwara et al. 2001). Our survey of IRAS galaxies is still ongoing, however, if we include the previous detection of the maser in NGC5793, the detection rate in our survey is significantly higher than those mentioned above. Although the survey is incomplete, its success seems to partially depend on the assumption that the masing can be enhanced by "warm'' dust emission heated by a nuclear radio source.

  \begin{figure} \par\includegraphics[width=8.8cm,clip]{h3059f1.ps} \end{figure} Figure 1: Spectra of water maser emission towards NGC6240 for different three epochs in 2001. The spectral channel spacing is 1.1 km s-1. $V_{\rm sys}$ is denoted by an arrow. Note that the flux density of the feature at $V_{\rm LSR} = 7565$ km s$^{-1}\;$clearly decreased at the last epoch  c).
Open with DEXTER

5 H $\mathsf{_2}$O maser in NGC6240

Figure 1 shows spectra of the H2O$\;$maser towards NGC6240 taken with the 100 m telescope at Effelsberg. The spectra show a prominent H2O$\;$maser feature with a narrow velocity-width (4.3 km s$^{-1}\;$[FWHM]) centered at $V_{\rm LSR} = 7565$ km s-1. The feature is redshifted by 261 km s$^{-1}\;$from the systemic velocity ( $V_{\rm sys}= 7304$ km s$^{-1}\;$with respect to the local standard of rest (LSR), based on HI observations; de Vaucouleurs et al. 1991). NGC6240 exhibits bright emission with a peak flux density of $\sim$40 mJy (Gaussian-fitted intensity is 36 mJy). Figure 2 shows a spectrum averaged over the two earlier observing epochs in Fig. 1 from which we estimate the apparent isotropic luminosity of the prominent feature alone to be $\sim$36 $L_{\odot}$ at a distance of 97 Mpc. In addition, we detect faint narrower (1.1 km s$^{-1}\;$[FWHM]) emission at $V_{\rm LSR} = 7609$ km s$^{-1}\;$(3$\sigma$ level detection). Including both emission features, the total luminosity of the maser is $\sim$40 $L_{\odot}$. Prior to our observations no H2O$\;$emission had been firmly detected (e.g. Claussen & Lo 1986), although tentative detections had been reported (Henkel et al. 1984; Braatz et al. 1994). The maser therefore appeared to be in a flaring state during our observing epochs in 2001. The peak flux density remained the same within the uncertainties of amplitude calibrations, at both early observing epochs. However, a third spectrum, taken on 12 July 2001 shows the strongest feature declining in flux to $\sim$27 mJy while the feature at 7609 km s$^{-1}\;$is more prominent (Fig. 1c).

  \begin{figure} \par\includegraphics[angle=-90,width=10cm,clip]{h3059f2.ps}\end{figure} Figure 2: Spectrum averaged over two observing epochs on 9 May and 14 June 2001. A downward arrow indicates the adopted systemic velocity of the galaxy, $V_{\rm LSR}=7304$ km s-1. Solid line indicates the result of 2-D Gaussian-fitting, assuming two Gaussian components in the observed velocity range. The fitted center velocities of each component are $V_{\rm LSR} = 7565$ km s-1( FWHM = 4.3 km s$^{-1}\;$) and 7609 km s$^{-1}\;$( FWHM = 1.1 km s-1) with peak flux densities of 35 mJy and 15 mJy, respectively.
Open with DEXTER


 

 
Table 2: NGC6240.
Distancea 97 Mpc
Systemic Velocity (21 cm HI)b,c $7304 \pm 9$ kms-1
                          (optical)b,c $7192 \pm 44$ kms-1
Inclination (optical) d 60 $\hbox{$^\circ$ }$
Optical classe LINER
$F_{\nu}$ (20 cm)f $386 \pm 19$ mJy
$F_{\nu}$ (100 $\mu$m)g $27.8 \pm 1.1$ Jy
$F_{\nu}$ (60 $\mu$m)g $22.7 \pm 0.9$ Jy
$F_{\nu}$ (25 $\mu$m)g $3.4 \pm 0.1$ Jy
$L_{\rm IR}$ (8-1000 $\mu$m)h $6.6\times10^{11}$ $L_{\odot}$
$L_{\rm X-ray}$ (2-10 keV)i $4\times10^{43}$- $6\times 10^{44}$ erg s-1


a Assuming H0= 75 kms-1 Mpc-1.
b Converted to LSR from helio centric velocity.
c de Vaucouleurs et al. (1991).
d Braatz et al. (1997).
e Heckman et al. (1987).
f VLA in B configuration (Colbert et al. 1994).
g Primary from NED.
h Sanders et al. (1988).
i Ikebe et al. (2000).


NGC6240 is categorized as an (ultra) luminous infrared galaxy ([U]LIRG) with a complex structure (Sanders et al. 1988) and is one of the most distant megamasers ever detected, rivaling the "Giga Maser'' in TXFS2226-184, which lies at a distance of 100 Mpc (Koekemoer et al. 1995). The basic characteristics of the galaxy are listed in Table 2. The large IR luminosity ( > 1010.8 $L_{\odot}$) is considered as having its origin in re-emission from a dust shell surrounding a hot core, suggesting the presence of an on-going starburst triggered by the merger of two galaxies (Sanders et al. 1988). However, recent hard X-ray measurements detected a strong iron-K emission line at $\sim$6.4 keV with an excess atomic hydrogen column density of $\sim$1022 cm-2 towards the nucleus, a phenomenon which appears to be common in the presence of an AGN (Iwasawa & Comastri 1998; Vignati et al. 1999; Ikebe et al. 2000). A 5 GHz high-resolution MERLIN continuum map (Beswick et al. 2001) revealed a double source with a separation of $\sim$1.5 $^{\prime\prime}$ ($\sim$700 pc). Beswick et al concluded that the radio continuum flux found in the galaxy is a combination of a starburst and a weak AGN induced by a merger event. A molecular gas concentration, which lies between the radio sources, was imaged in CO emission on 100 parsec-scales (Bryant & Scoville 1999; Tacconi et al. 1999; Tecza et al. 2000). According to Tacconi et al. (1999), the CO emission is settling down into the two nuclei (which correspond to the double peaks in the MERLIN map) and will likely form a central thin disk in the final stage of evolution.

6 What is the maser in NGC6240?

The detection of H2O$\;$masers from NGC6240, which is exhibiting starburst activity, raises questions concerning the nuclear activity of the galaxy. The major energy source is considered to be a nuclear starburst (e.g., Tecza et al. 2000) since the ratio of the hard X-ray luminosity to the IR luminosity is small, $\sim$0.01-0.1 (Table 2). Such a ratio suggests that the hard X-ray source (i.e. the AGN) is not dominant (e.g., Ikebe et al. 2000). What is the energy source that gives rise to the water maser in NGC6240?

H2O$\;$megamasers are often associated with a compact continuum source such as a radio core or jet. In the nuclear region of NGC6240 the two radio sources at 5, 8.4 and 15 GHz remain unresolved at $\sim$0.1 arcsec resolution. Each source has a brightness temperature upper limit ($T_{\rm b}$) of ${\sim} 10^4$-105 (K) (Colbert et al. 1994; Beswick et al. 2001). (There has been no reported detection of milliarcsecond-scale structure in the galaxy.)

Tacconi et al. (1999) imaged the CO(2-1) emission peak between the two radio sources with 0.7 $^{\prime\prime}$ resolution, the spectral profiles show wide line-widths spanning $\pm$500 km s$^{-1}\;$ with respect to the systemic velocity of 7339 km s-1. A similar result was obtained in HCN(1-0) emission, which is a more appropriate tracer of the warm dense molecular gas that can produce 22 GHz H2O$\;$maser emission. Both CO and HCN cover the velocities of the maser but there are no distinct velocity peaks corresponding to those of the maser, weakening the case for an association between the CO/HCN gas and the masing cloud.

The location of the H2O$\;$maser of NGC6240 is of great interest. MERLIN high-resolution HI absorption observations show that there exist velocity gradients against the two unresolved radio sources. The HI gas in front of the two sources differ in their velocity centroids by ${\sim} 150$ km s-1; the gradients range from $V_{\rm LSR} = 7100$-7200 km s$^{-1}\;$ and 7250-7350 km s$^{-1}\;$respectively (Beswick et al. 2001). However, there is no significant detection of HI gas at the H2O$\;$maser velocities of $V_{\rm LSR} = 7565$ and 7609 km s-1, implying the masers are not associated with the observed HI absorbing gas. It is thus unlikely that the maser emission is associated with these two radio sources, which are widely accepted as being supernova remnants. This is consistent with the fact that total maser luminosity, which we estimate to be $\sim$40 $L_{\odot}$ is an order of magnitude larger than those of extragalactic H2O$\;$masers outside AGN (0.1-1.0 $L_{\odot}$), i.e. kilomasers that are associated with star forming regions in nearby galaxies (Claussen et al. 1984; Ho et al. 1987; Greenhill et al. 1990). Consequently, it is likely that the maser in NGC6240 is directly associated with nuclear activity, the presence of which was probed by the hard X-ray measurements.

Given the detection of the features redshifted by 260-300 km s$^{-1}\;$from $V_{\rm sys}$ and the narrowness of each line, it is reasonable to assume that the maser might originate from a well-defined region where the masing gas is receding relative to the molecular gas disk observed on 100 pc-scales. One can also speculate on the possible existence of an edge-on molecular gas disk rotating around a super massive object by analogy with other megamasers (Moran et al. 1999). Such a hypothesis is supported by the nature of the observed spectrum of the H2O$\;$maser towards NGC 6240 and its similarity to that of the so-called disk masers. If such a scenario was true, the H2O$\;$maser should lie in the dynamical center of the galaxy, where the observed dense molecular gas will eventually settle down between the two radio sources and form a disk.

NGC6240 has often been compared with the prototype ULIRG Arp220 (IC4553) from which hard X-ray radiation or other evidence of AGN activity has never been incontrovertibly detected (Kii et al. 1997; Genzel et al. 1998; Smith et al. 1998 and references therein). The galaxy is a merging system showing a large concentration of dense molecular gas between twin stellar/starburst nuclei (Scoville et al. 1997). H2O$\;$maser emission has not been detected in Arp220 (Braatz et al. 1997) however, it is the prototype of the class of galaxies that exhibit OH megamaser emission. Recent surveys with Arecibo have revealed many new OH megamaser sources bringing the total number of detections to almost 100 (Darling & Giovanelli 2001). Most of the OH megamaser detections are in galaxies that are classified as [U]LIRGs, and all [U]LIRGs appear to be merging systems. The most remarkable thing is that H2O$\;$nuclear megamasers and OH megamasers are mutually exclusive: No galaxy observed to date contains both an OH megamaser and a nuclear H2O$\;$megamaser. NGC6240 does not show OH emission, however OH gas is seen in intense absorption (e.g. Baan et al. 1998). This is similar to the case of the OH absorption towards the nucleus of the megamaser NGC5793 (Hagiwara et al. 2000). Until now there have been no detections of H2O$\;$masers towards merging systems, NGC6240 appears to break the mould.

The presence of strong H2O$\;$maser emission and the iron-K line in NGC6240 might mean that the galaxy is experiencing a different phase of merging from that of Arp220 i.e. compact single disk formation surrounding a "central engine'', thereby placing it the final evolution stage in galaxy-galaxy merging.

7 Conclusions

We have made the first detection of H2O$\;$megamaser emission from a [U]LIRG/merger galaxy. We have detected emission in the LINER galaxy NGC6240, the maser features are observed to be by 260-300 km s$^{-1}\;$redshifted with respect to the systemic velocity and to have a luminosity of ($\sim$40 $L_{\odot}$). The maser could arise from a dense circumnuclear molecular cloud on a ${\sim} 100$ parsec scale or from a spatially compact maser disk inferred from the presence of the high velocity maser feature(s). The interpretation of the observed maser is not unique, though we favour the latter case. Interferometric observations of the new maser will provide an opportunity to investigate the kinematics of the ongoing merger.

Acknowledgements

YH would like to thank H. Imai for his assistance with a part of the observations. YH also appreciates R. Kawabe for helpful discussion on the source selection, and A. W. Sherwood and Chr. Henkel for able advice during the data analysis. We acknowledge M. J. Claussen for his suggestions on the manuscript. 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.

References

 
Copyright ESO 2002