Major impact from a minor merger

Major impact from a minor merger

The extraordinary hot molecular gas flow in the Eye of the NGC 4194 Medusa galaxy 1
Key Words.:
galaxies: evolution – galaxies: individual: NGC 4194 – galaxies: starburst – galaxies: active – radio lines: ISM – ISM: molecules


Context:Minor mergers are important processes contributing significantly to how galaxies evolve across the age of the Universe. Their impact on the growth of supermassive black holes (SMBHs) and star formation is profound – about half of the star formation activity in the local Universe is the result of minor mergers.

Aims:The detailed study of dense molecular gas in galaxies provides an important test of the validity of the relation between star formation rate (SFR) and HCN luminosity on different galactic scales – from whole galaxies to giant molecular clouds in their molecular gas-rich centers.

Methods:We use observations of HCN and HCO 10 with NOEMA and of CO 32 with the SMA to study the properties of the dense molecular gas in the Medusa merger (NGC 4194) at 1 resolution. In particular, we compare the distribution of these dense gas tracers with CO 21 high-resolution maps in the Medusa merger. To characterise gas properties, we calculate the brightness temperature ratios between the three tracers and use them in conjunction with a non-LTE radiative line transfer model.

Results:The gas represented by HCN and HCO 10, and CO 32 does not occupy the same structures as the less dense gas associated with the lower- CO emission. Interestingly, the only emission from dense gas is detected in a 200 pc region within the “Eye of the Medusa”, an asymmetric 500 pc off-nuclear concentration of molecular gas. Surprisingly, no HCN or HCO is detected for the extended starburst of the Medusa merger. Additionally, there is only little HCN or HCO associated with the AGN. The CO 32/21 brightness temperature ratio inside “the Eye” is 2.5 – the highest ratio found so far, and implying optically thin CO emission. The CO 21/HCN 10 (9.8) and CO 21/HCO 10 (7.9) ratios show that the dense gas filling factor must be relatively high in the central region, consistent with the elevated CO 31/21 ratio.

Conclusions:The line ratios reveal an extreme, fragmented molecular cloud population inside “the Eye” with large bulk temperatures (T  300 K) and high gas densities (n(H)  10 cm). This is very different from the cool, self-gravitating structures of giant molecular clouds normally found in the disks of galaxies. “The Eye of the Medusa” is found at an interface between a large-scale minor axis inflow and the central region of the Medusa. Hence, the extreme conditions inside “the Eye” may be the result of the radiative and mechanical feedback from a deeply embedded, young and massive super star cluster, formed due to the gas pile-up at the intersection. Alternatively, shocks from the inflowing gas entering the central region of the Medusa may be strong enough to shock and fragment the gas. For both scenarios, however, it appears that the HCN and HCO dense gas tracers are not probing star formation, but instead a post-starburst and/or shocked ISM that is too hot and fragmented to form new stars. Thus, caution is advised in linking the detection of emission from dense gas tracers to evidence of ongoing or imminent star formation.

1 Introduction

Figure 1: Left: Overlay of the high resolution CO 21 emission contours on top of an HST WFPC2 F606W filter image (from König et al. 2014). The insert shows a zoom into the 3 3 surrounding the “Eye of the Medusa”. Contours start at 2 and are spaced in steps of 2 (1 0.7 Jy beam km s). The locations of the most important dust lanes are indicated by grey and white curves. Center: Distribution of the integrated intensity emission of the 3 mm continuum (blue contours) compared to the CO 21 emission (grey contours and background). The 3 mm continuum contours start at 5 and are spaced in steps of 3 (1 50 Jy beam). The main 3 mm continuum emission peak is located at the center of the “Eye of the Medusa” (E). Right: 1 mm continuum emission on top of the CO 21 emission. Contours are at 3 and 6 (1 0.5 mJy beam). The emission peaks south of the AGN position, close to the center of “the Eye”. North is up, east to the left. The position of the 1.4 GHz continuum peak is marked by a white cross (Beswick et al. 2005). Beam sizes are 0.6 0.5 for CO 21, 0.9 0.7 for the 3 mm continuum, and 1.9 1.8 for the 1 mm continuum.

Galaxy evolution is a fundamental part of the overall evolution of the Universe. Its sphere of influence extends from the large scales dominated by dark matter, down to the small scales ruled by dissipative baryons that can form stars and grow supermassive black holes (SMBHs, Shlosman 2013). Interactions and mergers are a known and efficient mechanism for galaxy growth (e.g., Sandage 1990, 2005; Kormendy & Cornell 2004). Minor mergers (unequal mass progenitors, mass ratios: 1:4) occur much more frequently than major mergers (equal mass progenitors, e.g., Hernquist & Mihos 1995; Kaviraj et al. 2009; Lotz et al. 2011). Their impact on the growth of SMBHs and star formation is profound – about half of the star formation activity in the local Universe is the result of minor mergers (Kaviraj 2014, 2016).
Schmidt (1959) was the first to systematically study the connection between gas density and star formation rate (SFR) in the Milky Way. Using H, CO 10 and HI, Kennicutt (1998) determined star formation rate surface densities and gas surface densities in a sample of normal spirals and starburst galaxies - confirming the results of Schmidt (1959). Gao & Solomon (2004a, b) found a similar, but tighter correlation between gas density and SFR when using HCN to study the properties of dense gas in relation to star formation in luminous infrared galaxies (LIRGs), ultra-luminous infrared galaxies (ULIRGs) and normal spiral galaxies.
The Medusa merger (NGC~4194, D = 39 Mpc, 1 = 189 pc Beswick et al. 2005) is a minor merger that harbours a region of highly efficient star formation in its inner 2 kpc – the star formation efficiency (SFE, 1.5  10 yr, König et al. 2014) rivals even that of well-known ULIRGs (Aalto & Hüttemeister 2000). Low- to intermediate-density gas spans the main body of NGC 4194 – from the tidal tail at 4.7 kpc radius, down to the central starburst (Aalto et al. 2001). High-resolution CO observations have shown that molecular gas is associated with the minor axis dust lane crossing the galaxy’s main body (Aalto & Hüttemeister 2000). The brightest CO emission is found in a striking, off-nuclear structure called the “Eye of the Medusa” (König et al. 2014). A fraction of the star formation in NGC 4194 is going on in young super star clusters (SSCs, 5-15 Myr, Bonatto et al. 1999; Pellerin & Robert 2007) with a kpc-scale distribution (Weistrop et al. 2004; Hancock et al. 2006).
Single-dish observations revealed a global CO-to-HCN 10 luminosity ratio of 25 for NGC 4194 within a 29 beam (Costagliola et al. 2011), a significantly higher value than for ULIRGs (6, e.g., Solomon et al. 1992; Curran et al. 2000), indicating that the average fraction of dense gas is significantly lower despite the similar extreme SFE.
Here we present a high-angular resolution study of the dense molecular gas in the Medusa merger, using HCN, HCO and CO emission as tracers thereof. Throughout the paper, we are concerned with pure rotational transitions of HCN, HCO, CO, CH, SiO, HCO between upper state  =  and lower state  =  that are labeled .
In Sect. 2 we describe the observations and how the data were reduced and analysed, in Sect. 3 we present the results, and in Sect. 4 we discuss their implications.

2 Observations

2.1 NOEMA observations

The HCN and HCO 10 observations of NGC~4194 were carried out with the Northern Extended Millimeter Array (NOEMA) in the frame of science verification on March 10, 2015. Data were taken with seven antennas in extended configuration, with baselines between 32 and 760 m. Thus, the observations are sensitive to scales smaller than 13.5. The phase center of the observations was located at =12:14:09.660 and =+54:31:35.85 – the 1.4 GHz radio continuum peak (J2000, Beswick et al. 2005). The 3 mm-band receivers were tuned to 87.039 GHz to cover the HCN and HCO 10 lines in the 3.6 GHz bandwidth of WideX. Also, the CH, HCO 10, and SiO 21 lines were accessible in the same tuning. The instrumental spectral resolution was 1.95 MHz (6.7 km s). For analysis purposes we smoothed the data to 10 km s, resulting in 1 rms noise levels per channel of 0.4 mJy beam for both HCN and HCO 10. The antenna system noise temperature ranged from 70 to 160 K. During the observations, different sources were observed for calibration purposes: MWC~349 and LkHa~101 as flux calibrators, 3C~84 as bandpass calibrator, and J1150+497 and J1259+516 as phase calibrators. We estimate flux calibration uncertainty of about 15-20%.
Data reduction and analysis were performed using the CLIC and MAPPING software packages within GILDAS2. Applying a natural weighting scheme led to a nearly circular beam size of 1.0.

Figure 2: Integrated intensity distributions of the HCN 10 (left, in magenta), HCO 10 (centre, in red) and CO 32 emission contours (right, in light blue) at the same spatial resolution (1.1 1.0) on top of the CO 21 high-resolution emission (background and grey contours, König et al. 2014). Contours start at 5 and are spaced in steps of 3 (1(HCN 10) 20.2 mJy beam km s, 1(HCO 10) 21.2 mJy beam km s, 1(CO 32) 1.6 Jy beam km s). The inset in each of the images shows the higher resolution data for each tracer (beam sizes: 1.0 0.8, 1.0 0.8, 0.7 0.7). North is up, east to the left. The position of the 1.4 GHz continuum peak is marked by a white cross (Beswick et al. 2005).

2.2 SMA observations

Co 32

On April 10, 2016, additional CO 32 observations were obtained with the Submillimeter Array (SMA) in its extended configuration with baselines between 44 and 226 m. The 345 GHz receivers, tuned to 342.935 GHz, were used in conjunction with the correlator in 4 GHz mode with 128 channels and a spectral resolution of 0.8 MHz (0.7 km s). We smoothed the data to a velocity resolution of 10.6 km s, which results in a 1 rms noise level per channel of 29.9 mJy beam. The observed data set is sensitive to scales smaller than 2.5 at the observing frequency. For calibration, Callisto and MWC349a (flux), J1924-292 (bandpass), and J1419+543 and J1153+495 (phase) were observed. The SMA data were calibrated using the dedicated MIR/IDL SMA reduction package. The estimated flux calibration uncertainty is of the order of 20%. The visibilities were converted into FITS format and transferred to the GILDAS/MAPPING package for further imaging. Similar to the NOEMA observations, the phase center is located at =12:14:09.660 and =+54:31:35.85. With natural weighting the resulting synthesized beam is 1.1 1.0 with a position angle of 81, the beam of the uniformly weighted data cube is 0.74 0.66 (PA 99).

1 mm continuum

Observations of the CO and CO 21 line transitions with the SubMillimeter Array (SMA) were taken in compact (April 08, 2010) and very extended configuration (February 21, 2010). We extracted the 1 mm continuum information from these data sets. The phase center of the image presented in this work is located at =12:14:09.660 and =+54:31:35.85. The resulting synthesized beam is 1.9 1.8 (PA 46). Imaging the compact configuration alone results in a synthesized beam of 3.4 3.3, PA 22. Baseline lengths range between 38 and 509 m. Thus these data are sensitive to emission originating from scales smaller than 4.2 at the observing frequency of 230 GHz. We estimated calibration uncertainties on all fluxes of 20%. For more details regarding data reduction, array combination and analysis, see König et al. (2014).

Table 1: Characterising properties of the interferometric observations.
Line Observatory Configuration B3 B4 5
[GHz] [m] [m] [arcsec]
CO 32 342.935 SMA Extended 44 226 4.1
CO 21 228.631 SMA Compact + Very Extended 38 509 7.1
HCO 10 88.451 NOEMA 7A-Special 32 760 21.8
HCN 10 87.899 NOEMA 7A-Special 32 760 22.0

3 Results

3.1 Continuum

3 mm continuum

The 3 mm continuum, as pictured in Fig. 1, shows three distinct, point-like sources in the center of NGC 4194 – one at the position of the AGN (A), one inside the “Eye” (E), and one in the western arm (W) of the CO 21 distribution (König et al. 2014). The highest peak flux is found in the emission peak inside the “Eye” (see Fig. 1). The flux density recovered from the three components is 5.6  0.4 mJy (within 3 contours): 2.2  0.4 mJy for the AGN position (A), 2.0  0.4 mJy inside the “Eye” (E), and 1.2  0.2 mJy in the western arm (W).

1 mm continuum

Fig. 1 shows the 1 mm continuum distribution in the central 2 kpc of NGC 4194. The emission peak is associated with the “Eye of the Medusa”, slightly south of the AGN position. The integrated intensity within 3 is 4.10.5 mJy at a central frequency of 230.5 GHz, when combining compact and very extended configuration. Imaging only the compact configuration yields an integrated intensity of 15.50.5 mJy. In contrast to the 3 mm continuum emission, the peak associated with “the Eye” is of more circular shape (Figs. 1, 5). Also, whereas the 3 mm continuum shows a secondary emission peak in the western arm, the 1 mm continuum does not. For a side-by-side comparison of the 3 mm and 1 mm continuum emission distributions degraded to the spatial resolution of the 1 mm observations see Fig. 5.

3.2 Line emission

Hcn & Hco 10

The distributions of the HCN and HCO 10 emission are quite similar to each other in NGC 4194 (Fig. 2): the main emission peak is located inside the “Eye of the Medusa” (“E”) where CO 21 emission shows a distinct minimum (Fig. 1, König et al. 2014). This location is also situated inside the minor axis dust lane crossing the main body of NGC 4194. A secondary peak is located in the western arm of the CO 21 distribution (“W”). A close look at the HCO channel map also reveals low-level emission (at a 3 level) associated with the AGN position (“A”).
The overall HCO 10 emission line is brighter than the corresponding line transition of HCN (cf. Fig. 4). Measuring the intensities in the naturally weighted map with 10 km s wide channels results in integrated fluxes of 1.0  0.1 Jy km s for the HCO (654  61 mJy km s in the “Eye”, 374  48 mJy km s in the western arm) and 0.8  0.1 Jy km s for HCN (515  53 mJy km s in the “Eye”, 299  44 mJy km s in the western arm), respectively.

Co 32

Like HCN and HCO 10, CO 32 has its main peak inside the “Eye” (“E”). Other emission regions are located in the western arm (“W”), at the AGN position (“A”) and east of the AGN in the dust lane. Except for the vicinity of the AGN, these regions are all deficient in CO 21 (Fig. 2). The total integrated intensity of all components combined, in the lower resolution cube, amounts to 158.0  7.7 Jy km s (93.8  4.9 Jy km s in “the Eye”, and 39.5  3.4 Jy km s in the western arm).

Additional emission lines

Other relevant emission lines in the 3.6 GHz wide band are CH  = 10, HCO  = 10, SiO  = 21. CH has been detected with an integrated flux of 350  103 mJy km s (see Fig. 4). Due to the non-detection of HCO 10 and SiO 21, only upper limits of 2.4 mJy km s and 2.5 mJy km s, respectively, are available for the fluxes of these line transitions. The limits in the integrated fluxes have been obtained from an area corresponding to the largest extension of the HCO emission, and within a velocity range of 85 km s of the respective line centers.

3.3 Interferometric line ratios

The determination of molecular line ratios is a very powerful method to explore the physical and chemical processes in the interstellar medium. However, one has to be aware of possible implications when using interferometric data. Indeed, an interferometer is sensitive to a limited range of spatial scales. Extended structures might be filtered out if the shortest array baselines are still too long. When comparing two lines, one must therefore ensure that the corresponding data are sampling similar spatial scales. Table 1 gives an overview of the properties for each observational setup. , the maximum recoverable scale, is defined as  0.6  / B (e.g., the ALMA Cycle 5 Technical Handbook), where B is the shortest projected baseline length. For our observations ranges between 4.1 for CO 32 and 22.0 for HCN 10, i.e., significantly larger than the dip of emission in CO 21 in “the Eye”. Additionally, we smooth each data set to the same beam size (1.0 1.0) for the line ratio determination, and we extract the integrated intensities from within the same region.
The  = CO 32/CO 21 brightness temperature ratio is large, 2.5, at position E, inside the “Eye of the Medusa” (see Table 2). The ratios were determined inside a circular 1 polygon. This is slightly larger than the area inside “ the Eye” where CO 21 is deficient. Thus, it is very likely that the inner walls of the “Eye of the Medusa” contribute to the measured CO 21 flux inside the polygon. Hence, the CO 32/CO 21 ratio we determined is a lower limit to inside “the Eye”. To measure more precisely, higher-resolution observations are needed.

4 Discussion

Line “Eye of the Medusa” Western arm
6\saveFN\Tb7\saveFN\Tbint peak \useFN\Tb8\saveFN\Tbpeak FWHM9\saveFN\FWHM \useFN\Tb\useFN\Tbint peak \useFN\Tb\useFN\Tbpeak FWHM\useFN\FWHM
[K km/s pc] [K] [km/s] [K km/s pc] [K] [km/s]
CO 32 892.3  6.3 13.82  0.42 74.7  3.1 504.4  3.3 7.48  0.27 66.2  2.7
CO 21 356.8  0.6 3.74  0.08 88.8  2.0 208.4  0.4 2.01  0.06 91.7  2.9
HCO 10 45.3  0.5 0.80  0.03 63.7  3.3 35.9  0.3 0.62  0.03 55.7  2.9
HCN 10 36.4  0.2 0.53  0.03 59.6  4.2 23.7  0.3 0.51  0.04 63.6  8.6
Table 2: Brightness temperatures and line widths in the “Eye of the Medusa” and the western arm.
Figure 3: Results of the RADEX analysis of the CO 32/CO 21, CO 32/HCN 10, and CO 32/HCO 10 brightness temperature ratios. The white curves and associated numbers indicate T ratios toward the “Eye of the Medusa” of 2 and 2.25 for CO 32/CO 21, 25 for CO 32/HCN 10, and 20 for CO 32/HCO 10, respectively. The range in H densities and kinetic temperatures are the same for each of the three plots. The CO column density (N(CO)/v) is 10 cm.

The color scheme indicates possible brightness temperature ratios for each combination of tracers.

4.1 Dense gas distribution

A strong correlation between SFR and integrated molecular line luminosity from dense gas is seen in most galaxies (e.g., Gao & Solomon 2004a, b). For the Medusa this would imply that due to its high SFR (FIR: 6-7 M yr – Aalto & Hüttemeister 2000, H: 10 M yr – Gallagher in prep.) the HCN luminosity should be high – that is not the case: On global galaxy scales, single-dish observations revealed a low HCN-to-CO 10 luminosity ratio similar to values found for GMCs in the galactic disks of M~31 and the Milky Way (Brouillet et al. 2005; Costagliola et al. 2011; Matsushita et al. 2015), indicating that the fraction of dense gas should be significantly lower than in ULIRGs despite the similarly extreme SFE. However, our high-angular resolution observations paint a different picture: the dense gas is not distributed throughout the starburst region in the central 2 kpc of the Medusa (e.g., Armus et al. 1990; Wynn-Williams & Becklin 1993; Prestwich et al. 1994; Aalto & Hüttemeister 2000). Instead, it is located in a compact region inside the “Eye of the Medusa” where the CO 21 emission shows a distinct minimum (Fig. 2), thus explaining the relative faintness on global scales. A comparison between the CO 21 emission and the young superstar clusters (SSCs) in the Medusa has shown that many young massive clusters are located away from the bulk of the CO (Weistrop et al. 2004; Hancock et al. 2006; König et al. 2014) - the same effect is visible for the dense gas, only in a more extreme fashion. So far, only one cluster has been identified in the vicinity of the “Eye” (Hancock et al. 2006). However, the relative astrometric uncertainty between the optical and mm data preclude a definite conclusion.
An in-detail kinematic comparison between the HCN, HCO 10 and CO 32, and the CO 21 emission in the “Eye of the Medusa” shows strong evidence for their different origins. A CO 21 spectrum obtained from the center of the “Eye” shows a FWHM line width of 90 km s. The HCN and HCO 10 spectra obtained from within the same area instead show a FWHM line width of 60 km s (see Table 2). The centroid velocities of the lines, however, do not differ significantly (i.e., by 10 km s). This may suggest that CO 21 is tracing a more extended, lower density component. The CO 21 intensity at high velocities is roughly the same in the northern and southern part of the “Eye”, while at lower velocities it is strongly dominated by emission in the North. The dense gas emission does not show such a behaviour – their centroids do not move at all. Thus, the evidence presented here points toward different origins of the dense gas and the CO 21 emission.

4.2 Origin and physical properties of the dense gas emission in the “Eye”

Excitation and line ratios

Previous studies, using single-dish as well as interferometric observations, mark the we observed for “the Eye” (2.5, see Sect. 3.3) as an extreme case: observations of a large number of luminous infrared galaxies show that values of greater than unity are rare. So far the highest found are at 1.9 (e.g., Greve et al. 2009; Papadopoulos et al. 2012). Interferometric observations show that even more powerful ULIRGs, like e.g., Arp~220 and NGC~6240, show 1 (e.g., Greve et al. 2009; Sliwa & Downes 2017). Thus, ratios measured in the “Eye of the Medusa” indicate truly exceptional properties of the dense gas.
together with brightness temperature ratios of CO 32 with HCN 10 and HCO 10 (Table 2) can help us to infer the gas properties in this extraordinary region. We perform a line excitation analysis using the radiative transfer code RADEX (van der Tak et al. 2007) by exploring a grid of parameters in kinetic temperature (50-500 K) and H density (10 to 10 cm) for a CO column density (N(CO)/v) between 10 and 10 cm. Fig. 3 shows the results of the radiative transfer modeling: high values (2.2) of can only be achieved for warm gas (T 300 K), H densities of a few 10 cm and CO column densities of 10 cm. Assuming that HCN, HCO 10 and CO 32 trace the same gas component, i.e., the same physical conditions in the gas, we tune the relative abundance ratios for these three molecules to match the line ratios for a given density. Adopting also similar beam-filling factors for each tracer, we estimate relative abundances for CO/HCN of 4000 and for CO/HCO of 9000. However, since we only have two CO transitions (21 and 32) and one transition (10) each of HCN and HCO, the range of possible H gas densities is not very well constraint. To change this, observations of additional transitions of HCN and/or HCO are necessary.

The nature of the dense gas clouds

As a first step and the simplest scenario, we assume that the dense gas emission originates from one large virialised cloud with a radius of 1 with n(H)  5  10 cm and T300 K. This results in an expected linewidth (FWHM) of the emission line of about 540 km s. The observed linewidths for all four lines, however, are less than 100 km s inside “the Eye” (see Table 2). Furthermore, such a structure would have a gas mass of about 10 M which is larger than the gas mass of most galaxies. Thus, the single-cloud, high-filling factor scenario is unrealistic. Alternatively, it’s possible that the emission is coming from a larger number of small, self-gravitating dense gas clumps. Assuming a typical clump size of 5 pc instead, the scale of dense molecular clouds (see e.g., Lizano & Shu 1987; Nguyen Luong et al. 2011), results in linewidths of about 10 km s. This scenario is a possibility if the filling factor in the gas is high. As a result of this exercise, it seems that the dense gas in the “Eye of the Medusa” is most likely fragmented into a large number of small filaments or clouds.

4.3 What is inside the “Eye of the Medusa”?

Spectral index

Continuum observations in the radio and mm regimes can provide hints towards the excitation mechanism for the dense gas emission. For the Medusa, high-resolution continuum data at 150 MHz (LOFAR, König et al. in prep.), 1.4, 5, 8.4, 15 and 22 GHz (eMERLIN, VLA, Beck et al. 2014, König et al. in prep.), and at 86, 230 and 345 GHz (this work, König et al. 2014) show spectral indices indicative of a mixture of non-thermal synchrotron and thermal free-free emission. These emission processes are usually associated with supernovae (SNe) and AGN activity (synchrotron) and young star formation (free-free, e.g., Beck et al. 2014; Schober et al. 2017, and references therein).
In an earlier paper, we have shown that it is possible that the energetic output from a large number of supernova explosions could have led to the observed shell-like morphology of the CO 21 emission in “the Eye of the Medusa” (König et al. 2014). FIR and X-ray observations hint towards the presence of low-level AGN activity at the dynamic center of the Medusa (e.g., Kaaret & Alonso-Herrero 2008; Bernard-Salas et al. 2009; Lehmer et al. 2010). The spatially resolved studies (Kaaret & Alonso-Herrero 2008; Lehmer et al. 2010) found the activity to be associated with the main 1.4 GHz radio continuum emission peak north of “the Eye” (Beswick et al. 2005). So far no indications for an AGN inside the “Eye of the Medusa” have been found. Thus, supernovae are the most likely contributor of the synchrotron emission in “the Eye”.
The free-free component contributing to the continuum emission could be due to the presence of young star clusters inside “the Eye”, which could be heavily obscured (Hancock et al. 2006). However, the faint mm continuum emission towards higher frequencies for “the Eye” (see Sect. 3.1) indicates that there is very little cold dust emission associated with the dense gas in this region. The high gas temperature (see Sect. 4.2.1) most likely leads to the dust being heated and thus its emission peak is shifted to shorter thermal infrared wavelengths. Observations at these wavelengths, that would provide a test of this hypothesis, are currently not available for the Medusa.
The mixed spectral index indicates that the age of the star formation in “the Eye” would be at least a few 10 yrs, since at this stage in massive stellar evolution the first supernovae occur.


Single-dish observations of the total molecular gas and the dense molecular gas on large scales have shown differences in the global properties in a range of activity types in galaxies (e.g., Solomon et al. 1997; Graciá-Carpio et al. 2008; Krips et al. 2008; Privon et al. 2015). Several studies have found intrinsic differences between the way dense molecular gas content is traced by e.g., HCN and HCO, and the distribution of low- CO transitions as tracers of the total gas properties: , the HCN-to-CO ratio, is small in quiescent and starburst galaxies (0.3, Gao & Solomon 2004b; Krips et al. 2007; Matsushita et al. 2015, and references therein) compared to galaxies with AGN activity (up to 1 or more, Usero et al. 2004; Krips et al. 2007; Matsushita et al. 2015, and references therein). On small scales can be significantly different from values found on galaxy scales (e.g., in NGC~1068, NGC~6951, and M~51 Matsushita et al. 2015; Krips et al. 2007; Kohno et al. 1996; Krips et al. 2011). Our high-resolution (1) HCN 10-to-CO 21 line ratio is 0.20 in the “Eye of the Medusa”. An interferometric CO 21-to-CO 10 line ratio map yields values of 0.9 in “the Eye” (König et al., in prep.). Taking the CO ratio into account, the HCN 10-to-CO 21 ratio translates to HCN 10-to-CO 10 of 0.1. This places “the Eye” in a regime close to what has been found LIRGs and ULIRGs (e.g., Gao et al. 2007; Usero et al. 2015).
It has also been shown that , the HCN-to-HCO line ratio, can be significantly higher close to AGN (nuclear) compared to regions dominated by starbursts, and regions of more quiescent star formation (e.g., Graciá-Carpio et al. 2006, 2008; Krips et al. 2008; Kohno et al. 2008; Davies et al. 2012), though it can also be high in young (pre-synchrotron) star forming regions. inside “the Eye” is approximately 0.9, which would place the “Eye of the Medusa” in the regime of starburst galaxies.

CN as a tracer of radiative feedback

The reasons for the enhancement of HCN and/or HCO relative to CO and/or each other are diverse: strong UV/X-ray radiation fields (e.g., Usero et al. 2004; Meijerink et al. 2006, 2007; Krips et al. 2007, 2011; Matsushita et al. 2015), energetic particles from AGN and/or jets, cosmic rays (e.g., Meijerink et al. 2006; Bayet et al. 2010, and references therein), young star formation (e.g., Pirogov 1999; Boonman et al. 2001; Lahuis et al. 2007, and references therein), IR-pumping of HCN (e.g., Carroll & Goldsmith 1981; Aalto et al. 1995; Matsushita et al. 2015), high-temperature gas chemistry. A molecule that can help differentiate between these scenarios is CN. This molecule can be formed from HCN as a result of intense UV radiation in photodissociation regions (PDRs, e.g., Aalto et al. 2002; Baan et al. 2008; Han et al. 2015). Costagliola et al. (2011) found that in the Medusa, globally CN 10 is found at higher intensity than HCN 10. If we assume that the CN and HCN emission originates from the same spatial region inside the “Eye of the Medusa”, the CN-to-HCN 10 brightness temperature ratio is 2.1. This might imply that a large part of the HCN was photodissociated into CN due to the intense UV radiation from nearby young massive stars.


What we know so far about the properties of the central ISM in the Medusa is that the gas inside “the Eye” is hot and dense, that there is very little cold dust, and that a mixture of synchrotron and free-free emission processes take places in this region. What exactly is going on in this region is not clearly determined, yet. So far, the evidence seems to point towards post-burst radiative feedback from young massive stars as the energetic source of the observed phenomena. In this scenario, HCN and HCO 10, and CO 32 are enhanced in the dense, hot gas that is associated with deeply embedded, fragmenting star-formation cores.
Alternatively however, the dense gas properties could also be the immediate result of shocks, in which the gas is compressed and heated. CO studies have shown that shock-excited molecular gas can reach temperatures and densities similar, or above, to the values we find for the “Eye of the Medusa” (e.g., NGC~1068, NGC~6240, NGC~7130, Hailey-Dunsheath et al. 2012; Meijerink et al. 2013; Pozzi et al. 2017).
To be able to distinguish between the two scenarios and better determine the small-scale structure of the dense gas in “the Eye”, further observations are necessary. Higher-resolution data sets of dense gas tracers, like the ones discussed in this paper, will help determine the degree of fragmentation in the “Eye” ISM, and to track the locations of shocks and/or clouds in which the massive stars are embedded. Observations of tracers such as SiO, CHOH at millimeter wavelengths, and NIR H will directly reveal the presence of shocks, whereas CN observations will reveal UV-photon dominated regions close to stars. Also, shocks do not heat dust as effectively as they heat gas (e.g., Meijerink et al. 2013). Thus a good test will be to determine the conditions in the warm dust at infrared wavelengths.

4.4 The importance of interaction-induced gas flows

Studies have shown that the global CO-to-HCN ratio does not necessarily reflect on the small-scale properties of the dense gas, i.e., a global value is not representative of the small-scale processes leading to the observed ratios (e.g., in NGC~1068, NGC~6951, and M~51 Matsushita et al. 2015; Krips et al. 2007; Kohno et al. 1996; Krips et al. 2011). For the Medusa, its relatively high global CO-to-HCN ratio implies HCN, and thus the dense gas, to be faint. This is puzzling however, especially when taking into account the Medusa’s high star formation efficiency. Following Kennicutt (1998) and Gao & Solomon (2004b, a), one would expect the known star-forming complexes to be associated with dense molecular gas. Our high-resolution observations show, however, that the dense gas is located in a very compact configuration inside the “Eye of the Medusa”, away from most of the optically visible on-going star formation and the bulk of the molecular gas. However, the 3 mm continuum associated with the dense gas inside “the Eye” yields a SFR of about 2.8 M yr (following Murphy et al. 2011). Thus, obscured star formation seems to be on-going in this region. An important factor contributing to this puzzle could be the presence of inflowing gas in this minor merger.
It has been shown that the inflow of molecular gas in galaxy mergers or interactions can lead to an increase in and/or the triggering of star formation (e.g., Meier et al. 2002; Whitmore 2007; Turner et al. 2015, and references therein). The Medusa is suspected to harbor a molecular gas inflow via its kpc-scale dust lane to the nucleus (Aalto et al. 2010; König et al. 2014), similar to what has been found in e.g., NGC 5253 (Turner et al. 2015). For NGC 5253 it has been suggested that it has experienced an encounter with another galaxy in its history. Observations of the dense molecular gas, traced by CO 32, in this dwarf galaxy led the authors to conclude that the increased star formation activity they find in a gas cloud previously only weakly detected in CO 21 is the result of gas inflow. Molecular gas is being transported along the minor axis dust lane, east of the nucleus, towards the galaxy center (the “streamer”). Associated with the CO 32-detection is an “extremely” dusty gas cloud that contains two very young stellar clusters of a few times 10 M each (Calzetti et al. 2015). The star formation inside these clusters was likely triggered as a result of the interaction of the inflowing gas with the prevailing ISM (Turner et al. 2015).
This scenario is a strong possibility for NGC 4194 as well. However, the affected region in the “Eye of the Medusa” (185 pc) is much larger (about 100 times the area) than what has been found in NGC 5253 (31 pc  11 pc) – enough space to host several star clusters. Also, in contrast to NGC 5253, very little cold dust emission is associated with the dense gas in “the Eye” (see Sect. 4.3.1). Furthermore, only a small amount of the dense gas emission in the Medusa is found in the dust lane associated with the gas inflow (see Figs. 1, 2). Whether these differences could be due to the dissimilar size scales in NGC 5253 and the Medusa, or a difference in the time scales of the processes induced by the gas inflow, is unclear.

5 Conclusions

In this paper, we studied the distribution and properties of the dense gas in the Medusa merger through observations of HCN and HCO 10 and CO 32. Surprisingly, all these tracers are located inside the “Eye of the Medusa”, where CO 21 emission shows a distinct minimum. The gas inside “the Eye” is hot (300 K) and dense (10 cm). We propose two possible scenarios to explain what could cause the extreme properties of the gas inside the “Eye of the Medusa”: 1) they are caused by shocks as the gas flow, that transports molecular gas via the dust lane to the center of the Medusa, collides with the central ISM inside “the Eye”, or 2) they are an effect of supernova-induced shocks and/or radiative feedback from embedded massive star clusters, where the star formation is fed by the gas flow. Additional higher resolution observations (0.5) of e.g., HCN and HCO 32, and CN 21, and H NIR observations will help us disentangle the energetic processes in the “Eye of the Medusa” to determine what caused the extreme properties in the dense gas.

We thank the referee for useful comments. A.A. acknowledges the support of the Swedish Research Council (Vetenskapsrådet) and the Swedish National Space Board (SNSB). The Submillimeter Array is a joint project between the Smithsonian Astrophysical Observatory and the Academia Sinica Institute of Astronomy and Astrophysics and is funded by the Smithsonian Institution and the Academia Sinica. This paper is based in part on data obtained with the International LOFAR Telescope (ILT) under project code LC7_006. LOFAR (van Haarlem et al. 2013) is the Low Frequency Array designed and constructed by ASTRON. It has observing, data processing, and data storage facilities in several countries, that are owned by various parties (each with their own funding sources), and that are collectively operated by the ILT foundation under a joint scientific policy. The ILT resources have benefitted from the following recent major funding sources: CNRS-INSU, Observatoire de Paris and Université d’Orléans, France; BMBF, MIWF-NRW, MPG, Germany; Science Foundation Ireland (SFI), Department of Business, Enterprise and Innovation (DBEI), Ireland; NWO, The Netherlands; The Science and Technology Facilities Council, UK. MERLIN/eMERLIN is a National Facility operated by the University of Manchester at Jodrell Bank Observatory on behalf of STFC. 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.

Appendix A WideX spectrum of the NOEMA observations

Figure 4: NOEMA spectrum of the WideX 3.6 GHz bandwidth (at 1.95 MHz ( 6.6km s) channel spacing) extracted from the naturally weighted data cube. The position of the HCO, HCN and CH 10 lines are indicated in colour. Note that due to hyperfine splitting, CH 10 shows two emission line components.

Appendix B Comparison of the 3 mm and 1 mm continuum images at similar resolution

Figure 5: 3 mm (left, blue contours) and 1 mm continuum emission (right, green contours) at similar spatial resolution on top of the CO 21 emission (grey contours and background). The 3 mm continuum contours start at 5 and are spaced in steps of 5 (1 31 Jy beam). Contours for the 1 mm continuum emission are at 3 and 6 (1 0.5 mJy beam). The main continuum emission peak is located south of the AGN position, close to the center of “the Eye”. However, the 3 mm continuum peak emission appears elongated; the 1 mm continuum is of more spherical shape. Also, whereas the 3 mm continuum shows a secondary emission peak in the western arm, the 1 mm continuum does not. North is up, east to the left. The beam sizes are 1.9 1.8 for both images.


  1. thanks: Based on observations carried out under project number I15AA001 with the IRAM NOEMA Interferometer. IRAM is supported by INSU/CNRS (France), MPG (Germany) and IGN (Spain).
  3. Shortest baseline in the data set.
  4. Longest baseline in the data set.
  5. Maximum recoverable scale. The interferometer in the given configuration is sensitive to structures smaller than this size scale.
  6. The brightness temperatures T were determined from resolution-matched (1) data cubes inside the same polygons of 1 size.
  7. is the brightness temperature integrated over the 1 polygon.
  8. peak T is the peak brightness temperature of the spectrum integrated over the 1 polygon.
  9. Obtained from a single Gaussian fit to the line.


  1. Aalto, S., Beswick, R., & Jütte, E. 2010, A&A, 522, A59
  2. Aalto, S., Booth, R. S., Black, J. H., & Johansson, L. E. B. 1995, A&A, 300, 369
  3. Aalto, S. & Hüttemeister, S. 2000, A&A, 362, 42
  4. Aalto, S., Hüttemeister, S., & Polatidis, A. G. 2001, A&A, 372, L29
  5. Aalto, S., Polatidis, A. G., Hüttemeister, S., & Curran, S. J. 2002, A&A, 381, 783
  6. Armus, L., Heckman, T. M., & Miley, G. K. 1990, ApJ, 364, 471
  7. Baan, W. A., Henkel, C., Loenen, A. F., Baudry, A., & Wiklind, T. 2008, A&A, 477, 747
  8. Bayet, E., Hartquist, T. W., Viti, S., Williams, D. A., & Bell, T. A. 2010, A&A, 521, A16
  9. Beck, S. C., Lacy, J., Turner, J., Greathouse, T., & Neff, S. 2014, ApJ, 787, 85
  10. Bernard-Salas, J., Spoon, H. W. W., Charmandaris, V., et al. 2009, ApJS, 184, 230
  11. Beswick, R. J., Aalto, S., Pedlar, A., & Hüttemeister, S. 2005, A&A, 444, 791
  12. Bonatto, C., Bica, E., Pastoriza, M. G., & Alloin, D. 1999, A&A, 343, 100
  13. Boonman, A. M. S., Stark, R., van der Tak, F. F. S., et al. 2001, ApJ, 553, L63
  14. Brouillet, N., Muller, S., Herpin, F., Braine, J., & Jacq, T. 2005, A&A, 429, 153
  15. Calzetti, D., Johnson, K. E., Adamo, A., et al. 2015, ApJ, 811, 75
  16. Carroll, T. J. & Goldsmith, P. F. 1981, ApJ, 245, 891
  17. Costagliola, F., Aalto, S., Rodriguez, M. I., et al. 2011, A&A, 528, A30
  18. Curran, S. J., Aalto, S., & Booth, R. S. 2000, A&AS, 141, 193
  19. Davies, R., Mark, D., & Sternberg, A. 2012, A&A, 537, A133
  20. Gao, Y., Carilli, C. L., Solomon, P. M., & Vanden Bout, P. A. 2007, ApJ, 660, L93
  21. Gao, Y. & Solomon, P. M. 2004a, ApJS, 152, 63
  22. Gao, Y. & Solomon, P. M. 2004b, ApJ, 606, 271
  23. Graciá-Carpio, J., García-Burillo, S., Planesas, P., & Colina, L. 2006, ApJ, 640, L135
  24. Graciá-Carpio, J., García-Burillo, S., Planesas, P., Fuente, A., & Usero, A. 2008, A&A, 479, 703
  25. Greve, T. R., Papadopoulos, P. P., Gao, Y., & Radford, S. J. E. 2009, ApJ, 692, 1432
  26. Hailey-Dunsheath, S., Sturm, E., Fischer, J., et al. 2012, ApJ, 755, 57
  27. Han, X. H., Zhou, J. J., Wang, J. Z., et al. 2015, A&A, 576, A131
  28. Hancock, M., Weistrop, D., Nelson, C. H., & Kaiser, M. E. 2006, AJ, 131, 282
  29. Hernquist, L. & Mihos, J. C. 1995, ApJ, 448, 41
  30. Kaaret, P. & Alonso-Herrero, A. 2008, ApJ, 682, 1020
  31. Kaviraj, S. 2014, MNRAS, 440, 2944
  32. Kaviraj, S. 2016, in IAU Symposium, Vol. 319, Galaxies at High Redshift and Their Evolution Over Cosmic Time, ed. S. Kaviraj, 130–136
  33. Kaviraj, S., Peirani, S., Khochfar, S., Silk, J., & Kay, S. 2009, MNRAS, 394, 1713
  34. Kennicutt, Jr., R. C. 1998, ApJ, 498, 541
  35. Kohno, K., Kawabe, R., Tosaki, T., & Okumura, S. K. 1996, ApJ, 461, L29
  36. Kohno, K., Nakanishi, K., Tosaki, T., et al. 2008, Ap&SS, 313, 279
  37. König, S., Aalto, S., Lindroos, L., et al. 2014, A&A, 569, A6
  38. Kormendy, J. & Cornell, M. E. 2004, in Astrophysics and Space Science Library, Vol. 319, Penetrating Bars Through Masks of Cosmic Dust, ed. D. L. Block, I. Puerari, K. C. Freeman, R. Groess, & E. K. Block, 261
  39. Krips, M., Martín, S., Eckart, A., et al. 2011, ApJ, 736, 37
  40. Krips, M., Neri, R., García-Burillo, S., et al. 2007, A&A, 468, L63
  41. Krips, M., Neri, R., García-Burillo, S., et al. 2008, ApJ, 677, 262
  42. Lahuis, F., Spoon, H. W. W., Tielens, A. G. G. M., et al. 2007, ApJ, 659, 296
  43. Lehmer, B. D., Alexander, D. M., Bauer, F. E., et al. 2010, ApJ, 724, 559
  44. Lizano, S. & Shu, F. H. 1987, in NATO ASIC Proc. 210: Physical Processes in Interstellar Clouds, ed. G. E. Morfill & M. Scholer, 173–193
  45. Lotz, J. M., Jonsson, P., Cox, T. J., et al. 2011, ApJ, 742, 103
  46. Matsushita, S., Trung, D.-V., Boone, F., et al. 2015, ApJ, 799, 26
  47. Meier, D. S., Turner, J. L., & Beck, S. C. 2002, AJ, 124, 877
  48. Meijerink, R., Kristensen, L. E., Weiß, A., et al. 2013, ApJ, 762, L16
  49. Meijerink, R., Spaans, M., & Israel, F. P. 2006, ApJ, 650, L103
  50. Meijerink, R., Spaans, M., & Israel, F. P. 2007, A&A, 461, 793
  51. Murphy, E. J., Condon, J. J., Schinnerer, E., et al. 2011, ApJ, 737, 67
  52. Nguyen Luong, Q., Motte, F., Schuller, F., et al. 2011, A&A, 529, A41
  53. Papadopoulos, P. P., van der Werf, P. P., Xilouris, E. M., et al. 2012, MNRAS, 426, 2601
  54. Pellerin, A. & Robert, C. 2007, MNRAS, 381, 228
  55. Pirogov, L. 1999, A&A, 348, 600
  56. Pozzi, F., Vallini, L., Vignali, C., et al. 2017, MNRAS, 470, L64
  57. Prestwich, A. H., Joseph, R. D., & Wright, G. S. 1994, ApJ, 422, 73
  58. Privon, G. C., Herrero-Illana, R., Evans, A. S., et al. 2015, ApJ, 814, 39
  59. Sandage, A. 1990, JRASC, 84, 70
  60. Sandage, A. 2005, ARA&A, 43, 581
  61. Schmidt, M. 1959, ApJ, 129, 243
  62. Schober, J., Schleicher, D. R. G., & Klessen, R. S. 2017, MNRAS, 468, 946
  63. Shlosman, I. 2013, Cosmological Evolution of Galaxies, ed. J. Falcón-Barroso & J. H. Knapen, 555
  64. Sliwa, K. & Downes, D. 2017, A&A, 604, A2
  65. Solomon, P. M., Downes, D., & Radford, S. J. E. 1992, ApJ, 387, L55
  66. Solomon, P. M., Downes, D., Radford, S. J. E., & Barrett, J. W. 1997, ApJ, 478, 144
  67. Turner, J. L., Beck, S. C., Benford, D. J., et al. 2015, Nature, 519, 331
  68. Usero, A., García-Burillo, S., Fuente, A., Martín-Pintado, J., & Rodríguez-Fernández, N. J. 2004, A&A, 419, 897
  69. Usero, A., Leroy, A. K., Walter, F., et al. 2015, AJ, 150, 115
  70. van der Tak, F. F. S., Black, J. H., Schöier, F. L., Jansen, D. J., & van Dishoeck, E. F. 2007, A&A, 468, 627
  71. van Haarlem, M. P., Wise, M. W., Gunst, A. W., et al. 2013, A&A, 556, A2
  72. Weistrop, D., Eggers, D., Hancock, M., et al. 2004, AJ, 127, 1360
  73. Whitmore, B. C. 2007, in IAU Symposium, Vol. 237, Triggered Star Formation in a Turbulent ISM, ed. B. G. Elmegreen & J. Palous, 222–229
  74. Wynn-Williams, C. G. & Becklin, E. E. 1993, ApJ, 412, 535
Comments 0
Request Comment
You are adding the first comment!
How to quickly get a good reply:
  • Give credit where it’s due by listing out the positive aspects of a paper before getting into which changes should be made.
  • Be specific in your critique, and provide supporting evidence with appropriate references to substantiate general statements.
  • Your comment should inspire ideas to flow and help the author improves the paper.

The better we are at sharing our knowledge with each other, the faster we move forward.
The feedback must be of minimum 40 characters and the title a minimum of 5 characters
Add comment
Loading ...
This is a comment super asjknd jkasnjk adsnkj
The feedback must be of minumum 40 characters
The feedback must be of minumum 40 characters

You are asking your first question!
How to quickly get a good answer:
  • Keep your question short and to the point
  • Check for grammar or spelling errors.
  • Phrase it like a question
Test description