AFTER THE INTERACTION: AN EFFICIENTLY STAR-FORMING MOLECULAR DISK IN NGC 5195

The 3.6–4.5–8.0 μm Spitzer (Werner et al. 2004) IRAC (Fazio et al. 2004) data that are shown in Figure 1 were obtained by downloading supermosaics from the Spitzer Heritage Archive.¹¹ These data were originally part of the Spitzer Infrared Nearby Galaxy Survey (Kennicutt et al. 2003). The observations were taken in 2004 May, with exposure times of 26.8 s and were processed through the Spitzer Science Center reduction pipeline, version S18.25.

Zoom In Zoom Out Reset image size

The Herschel Photo-detecting Array Camera and Spectrometer (PACS; Poglitsch et al. 2010) data were observed in 2009 December in the "blue channel" (70 and 160 μm) and were downloaded from the Herschel Science Archive (PI: C. Wilson, KPGT_cwilso01; see: Parkin et al. 2013 for the 70 μm image). The Level-1 data were reduced using the scanamorphos (Roussel 2013) package that accurately fits the low-frequency (1/f) noise. Typical sky background measurements in the final maps were 0.13 and 0.25 mJy for the 70 and 160 μm bands, respectively.

The optical spectrum used below was obtained through the Sloan Digital Sky Survey (SDSS), which surveyed galaxies using a dedicated 2.5 m telescope at Apache Point Observatory. The reduced and calibrated spectral data were obtained from SDSS Data Release 12 (Alam et al. 2015), without any additional modifications applied.

The 1.4 GHz data for NGC 5195, taken as part of the Westerbork SINGS survey (Braun et al. 2007) were downloaded for analysis. Observations of the M51 complex were taken on 2003 June 25 and Nov 23, with a beamsize of 17'' × 125. Because the 1.4 GHz emission associated with NGC 5195 has a complex distribution, we measured the 1.4 GHz flux density from the SINGs image by summing the flux within an aperture matching the extent of the CO emission using the Common Astronomy Software Application¹² viewer (McMullin et al. 2007).

HCN(1–0) and HCO⁺(1–0) were observed at the Onsala Space Observatory, Sweden, on 2016 March 3–4, April 15 and 26. The SIS 3 mm receiver was connected to the Omnisys A spectrometer, which gives a bandwidth 4 GHz wide with dual polarization. The observations were performed with a dual-beam switch mode with a throw of $11^{\prime}$ in azimuth. The half power-beam width of the telescope at the rest frequency (88.9 GHz) was $43^{\prime\prime}$ and the pointing accuracy was better than $3^{\prime\prime}$. The focus was checked each day on bright quasars and the corrections were ∼1 mm. The average system temperature was around 150 K, and the opacity 0.2.

We converted the data from antenna temperature to main-beam temperature correcting by a beam efficiency of 0.53. The spectrum was averaged to a final velocity resolution of 40 km s⁻¹ (12 MHz), for which the achieved rms is 0.145 mK. We fitted a baseline of order two to subtract the continuum.

The spectra of the HCN and HCO⁺ lines are shown in Figure 2. Using the K per Jy factor of 0.07999, we measure a total HCN(1–0) line flux of 3.25 ± 0.27 Jy km s⁻², with velocity width of 219 km s⁻¹. We also measure a HCO⁺(1–0) total line flux of 2.09 ± 0.24 Jy km s⁻² with a velocity width of 193 km s⁻¹.

Zoom In Zoom Out Reset image size

We measure an HCN/HCO⁺ J = 1–0 line ratio of 2. Reproducing HCN(1–0)/HCO⁺(1–0) of 2 with n = 10⁴ cm⁻³ and T_K = 65 K requires a higher HCN than HCO⁺ abundance. As long as the emission is not very optically thick the the lower critical density of HCO⁺ would render its J = 1–0 transition more luminous than that of HCN (for the same abundance). For the given physical conditions above the HCN(1–0)/HCO⁺(1–0) ratio should be about 0.5 for the same abundance (and for moderate optical depths). To obtain an HCN(1–0)/HCO⁺(1–0) of 2 for n = 10⁴ cm⁻³ and T_K = 65 K, requires the HCN/HCO⁺ abundance ratio to be about 10.

We present observations of NGC 5195 that were pulled from the CARMA archive, as well as original observations. The archival portion of our NGC 5195 data was observed in ¹²CO(1–0) with CARMA between 2012 May and December in two different configurations: C-array (1'', 30–350 m baselines), D array (3'', 11–150 m baselines) by the CARMA & Nobeyama Nearby Galaxies (Donovan Meyer et al. 2013) survey. We followed up these archival observations in 2014 June with observations of NGC 5195 in E-array (8'', 8–66 m baselines), to better detect some of the more diffuse gas that might be present. ¹²CO(1–0),¹³CO(1–0), CN(1${}_{\mathrm{0,2}}$–0${}_{\mathrm{0,1}}$) and CS(2–1) were observed simultaneously in this set of E-array observations, utilizing the upgraded correlator. The observation properties for each molecular line detected by CARMA in NGC 5195 are listed in Table 2. All data were reduced and imaged using the Multichannel Image Reconstruction, Interactive Analysis and Display software (miriad; Sault et al. 1995).

Table 2.  Derived Molecular Properties

		12CO(1–0)	13CO(1–0)	CN(1–0)	CS(2–1)
${\nu }_{\mathrm{rest}}$	(GHz)	115.2712	110.2014	113.4910	97.9810
${\theta }_{\mathrm{beam}}$	('')	5.0 × 4.4	7.4 × 5.9	7.2 × 5.8	8.3 × 6.7
K Jy−1	⋯	4.126	2.290	2.291	2.290
${v}_{\mathrm{range}}$	(km s−1)	463–799	448–729	298–740	473–674
${\rm{\Delta }}v$	(km s−1)	10	20	40	50
rms	(mJy bm−1)	12.0	3.3	4.7	4.0
${I}_{\mathrm{peak}}$	(K km s−1)	243.8 ± 15.1	23.9 ± 2.3	19.6 ± 2.3	5.7 ± 1.1
Mom0 area	(arcseconds)a	1489	763	365	360
	(kpc2)	3.29	1.69	0.81	0.80
rms/channel	(mJy)	97.3	13.0	13.2	9.6
Line flux	(Jy km s−1)	243.8 ± 5.2	22.7 ± 1.0	12.4 ± 1.4	3.9 ± 1.0
S/N (Line)	⋯	46.9	22.7	8.9	3.9
${L}_{\mathrm{line}}$	(${L}_{\odot }$)	2859 ± 61	255 ± 11	143 ± 16	39 ± 10

Note.

^aarcsec².

Download table as:  ASCIITypeset image

The primary beam has a diameter of $2^{\prime}$ at CO(1–0), which covers all emission from NGC 5195. M51a itself does appear within the primary beam, but is easily separable from NGC 5195 (see Figure 1(a)). All observations used a long integration on a bright quasar to calibrate the passband, and alternated integrations between a gain calibrator (J1419543) and NGC 5195. We used the miriad task uvlin to separate out continuum emission from the line emission, and we estimate the continuum flux of NGC 5195 to be 1.21 ± 0.27 mJy, centered on 106 GHz, from a CARMA image with a 76 × 61 beam.¹³ The calibration and data reduction steps were followed identically to those in Alatalo et al. (2013) to construct channel maps, and moment maps.

Channel maps are shown in Figures 3–6, with the individual gas channels overlaid on a 8 μm nonstellar Spitzer IRAC image. To create the underlying Spitzer image, the 8.0 μm IRAC4 data was corrected for the stellar contribution by subtracting the 3.6 μm IRAC1 image, normalized to the expected stellar continuum at 8 μm (0.232; Helou et al. 2004). ¹²CO, ¹³CO, and CN(1–0) are detected. The CN(1–0) channel map includes blueshifted emission compared to the line peak (≈250 km s⁻¹ away, at 340 km s⁻¹), which we have attributed to M51a and do not consider in subsequent analysis. CS(2–1) is weak, but is still detected with a signal-to-noise ratio (S/N) of 3.9. A tentative line, identified as CH₃CN(6_k–5_k), seems to be present in the ¹³CO spectrum outside of the velocity range that would be inhabited by ¹³CO in M51a, and has a line flux of 1.27 ± 0.52 (only detected with 2.4σ significance), but requires deeper observations to confirm whether it is real.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Figures 1 and 7 display the integrated intensity (moment0), and mean velocity (moment1) maps, and Figure 8 shows the integrated spectra of the ¹²CO, ¹³CO, CN(1–0), and CS(2–1) within NGC 5195, Figure 9 shows the position–velocity (PV) diagrams across the specified slices of the ¹²CO, ¹³CO, and CN detections (the CS(2–1) being too low in S/N and compact to display). Of note, while the ¹²CO and ¹³CO share the same kinematic axis (of 30°), the CN (with an angle of 0°) has a different kinematic major axis from the other gas tracers and is thus considered kinematically misaligned according to Davis et al. (2011).

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

The integrated spectrum for each line was constructed using the moment0 map to create a clip-mask and integrating the flux within the moment0-defined (unmasked) aperture in each channel of the data cube. This was done separately for each tracer. The rms noise was then calculated as the standard deviation of all pixels in the cube outside of the moment0-aperture per channel and is listed in Table 2. An additional 30% correction was also added in quadrature to the rms noise to account for the oversampling of the maps (see: Alatalo et al. 2015b for details). The rms noise per channel for the spectrum was then calculated by multiplying the rms of the entire data cube by the square root of the total number of beams in the moment0-aperture.

The line luminosity for each molecule was calculated using Solomon & Vanden Bout (2005):

$$\begin{eqnarray}&&{L}_{\mathrm{line}}=1.04\times {10}^{-3}\,{S}_{\mathrm{line}}{\rm{\Delta }}v\,{\nu }_{\mathrm{rest}}{(1+z)}^{-1}{D}_{L}^{2},\end{eqnarray} \tag{ 1 }$$

where ${S}_{\mathrm{line}}{\rm{\Delta }}v$ is the line flux in [Jy km s⁻¹], ${\nu }_{\mathrm{rest}}$ is the rest frequency of the line in [GHz], z is the redshift of NGC 5195 (calculated using a systemic velocity of 634 km s⁻¹, the CO(1–0) spectrum), and D_L is the luminosity distance, defined as 9.9 Mpc in Section 1. The resultant line fluxes are listed in solar luminosities in Table 2.

Kohno et al. (2002) observed ¹²CO(1–0) and HCN(1–0) in NGC 5195 with the NRO 45 m single dish, and imaged the ¹²CO(1–0) with the NMA Interferometer. The new OSO HCN(1–0) observations are consistent with the relatively faint HCN emission found by Kohno et al. (2002). Using the 45 m, Kohno et al. (2002) find a total line flux of 199 Jy km s⁻¹ (using a 2.4 Jy K⁻¹ conversion; Ueda et al. 2014), in conflict with their resultant interferometric flux of 340 Jy km s⁻¹. We have measured a ¹²CO(1–0) line flux of 243.8 ± 5.2 Jy km s⁻¹, falling right between the two estimates of those authors (34% larger than the 45 m, and 27% lower than the NMA line fluxes). Were some of the M51a emission to be included in the map, it would result in an overestimation of the CO(1–0) line flux for NGC 5195, although Kohno et al. (2002) make note of the low-velocity component likely being from the spiral arm of NGC 5194 (displayed as gray contours in the first nine channels in Figure 3). If we also account for the 20% absolute flux uncertainties associated with the 3 mm flux calibration (for both the NMA and the CARMA observations), we consider ourselves to be in reasonable agreement with Kohno et al. (2002).

Sage (1990) presented ¹²CO(2–1) and ¹³CO(1–0) observations of NGC 5195 from the NRAO 12 m, measuring a ¹³CO line flux of 12.5 ± 3.3 Jy km s⁻¹ (using the Jy/K = 30.4 of the NRAO 12m; Ueda et al. 2014). We have detected nearly twice the total ¹³CO flux of Sage (1990), although that is likely due to S/N (our detection has six times higher S/N than Sage 1990). It is also possible that pointing errors and calibration uncertainties might also play a part. Matsushita et al. (2010) used the NRO 45 m to measure the ¹³CO(1–0) of NGC 5195 to be 28.8 Jy km s⁻¹, in much better agreement (within 22%) with our ¹³CO measurement. Overall, our observations are within reasonable agreement with those done previously.

We calculate the molecular mass expected in NGC 5195 using the ¹²CO(1–0) luminosity. Assuming that NGC 5195 is similar to the Milky Way, where we assume that molecular gas is distributed within giant molecular clouds, the conversion between the CO luminosity and the H₂ mass (from Bolatto et al. 2013) is:

$$\begin{eqnarray}&&\displaystyle \frac{{M}_{{{\rm{H}}}_{2}}}{{L}_{\mathrm{CO}}}=8.75\times {10}^{4}\,\displaystyle \frac{{M}_{\odot }}{{L}_{\odot }},\end{eqnarray} \tag{ 2 }$$

which corresponds to a ${X}_{\mathrm{CO}}$ conversion factor of 2 × 10²⁰ cm⁻²(K km s⁻¹)⁻¹, considered the standard, holding both for the Milky Way and most normal nearby galaxies. Using this conversion, and ${L}_{\mathrm{CO}}$ ¹⁴ from Table 2, we find a total H₂ mass of (2.74 $\pm \,0.06)\,\times$ 10⁸ ${M}_{\odot }$.¹⁵

To calculate the H₂ surface density, ${{\rm{\Sigma }}}_{{{\rm{H}}}_{2}}$, we use the area subtended by the ¹²CO(1–0) clipped moment map (listed in Table 2), converted from arcseconds to the physical scale of the galaxy to be 3.15 kpc², which corresponds to a mean H₂ surface density, $\langle {{\rm{\Sigma }}}_{{{\rm{H}}}_{2}}\rangle$ = 87 ± 2 ${M}_{\odot }$ pc⁻². To calculate the peak ${{\rm{\Sigma }}}_{{{\rm{H}}}_{2}}$, we convert the peak line intensity to the H₂ surface density, assuming the conversion from Bolatto et al. (2013):

$$\begin{eqnarray}&&\displaystyle \frac{{{\rm{\Sigma }}}_{{{\rm{H}}}_{2}}}{{M}_{\odot }\,{\mathrm{pc}}^{-2}}=\displaystyle \frac{3.16\,{I}_{\mathrm{peak}}}{{\rm{K}}\,\mathrm{km}\,{{\rm{s}}}^{-1}},\end{eqnarray} \tag{ 3 }$$

where ${I}_{\mathrm{peak}}$ represents the peak intensity of the ¹²CO map, resulting in ${{\rm{\Sigma }}}_{{{\rm{H}}}_{2},\mathrm{peak}}$ = 810 ± 50 ${M}_{\odot }$ pc⁻², made over an area of 0.05 kpc² (one beam), an order of magnitude larger than $\langle {{\rm{\Sigma }}}_{{{\rm{H}}}_{2}}\rangle$, and consistent with the range seen in nearby disk galaxies (Bigiel et al. 2008; Leroy et al. 2008).

Sage (1990) measured the ¹²CO and ¹³CO in NGC 5195 (discussed in Section 2.4), finding a ¹²CO/¹³CO flux ratio (hereafter, ${{ \mathcal R }}_{12/13}$) to be ≈26. This high ratio is often seen in hot dust hosts and ULIRGs (Aalto et al. 1995), suggestive of a hot, disrupted molecular disk in which the ¹²CO emission includes a large optically thin contribution. The CARMA-derived ¹²CO and ¹³CO fluxes shown in Table 2 update ${{ \mathcal R }}_{12/13}$ in NGC 5195 to be 11.4 ± 0.5, which is more consistent with normal star-forming galaxies as well as field early-type galaxies (Alatalo et al. 2015b) than interacting galaxies. The revised ${{ \mathcal R }}_{12/13}$ in NGC 5195 seems to suggest that the dynamical state of the gas is not as disrupted as originally thought, having had time to settle since the last interaction.

The CN abundance can be enhanced either by UV fields (for example as a photo-dissociation product of HCN in the envelopes of molecular clouds) or by X-rays that can penetrate further into the inner, denser regions. Hence, CN is a molecular tracer of photo-dissociated regions (PDRs) and X-ray dissociated regions (XDRs), depending on the environment (e.g., Lepp & Dalgarno 1996; Rodriguez-Franco et al. 1998). The strong PAH, IR, and Hα emission in NGC 5195 (Boulade et al. 1996; Greenawalt et al. 1998 and Figure 1) suggest that the CN is associated with the old starburst caused by the encounter with M51a and may be associated with PDRs.

To explore the possibility that CN could be tracing XDRs, we compared our CN intensity maps with the X-ray emission from Schlegel et al. (2016), where Chandra images show two recent outbursts coming out from the supermassive black hole in NGC 5195. A comparison of the CN and X-ray images shows no obvious spatial correlation between them.

In a barred galaxy potential, dynamical models predict two kinds of orbits: (1) x1 orbits oriented parallel to the major axis of the bar that support the bar and (2) x2 orbits aligned perpendicular to the bar and located closer to a galaxy's nucleus. Gas in self-intersecting x1 orbits tends to shock and lose energy, eventually changing orientation and moving inward to the lower-energy x2 orbits. For a review, we refer readers to Alloin (2006). The moment maps in Figure 7 show that CN arises from the inner regions of the galaxy with isovelocity contours at a different angle than those of ¹²CO and ¹³CO. The latter seem to trace a more extended gas distribution, possibly associated with the x1 orbits of the large-scale bar (Kohno et al. 2002), while CN could be tracing the perpendicular x2 orbits of the inner region. It is possible that this difference in alignment is due to resolution effects, but given the high S/N CN channels seen in Figure 5, we believe the detection of differing alignments is robust.

This would be a similar scenario to that observed in the central 100 pc of M 82, where the CN abundance (among other PDR tracers) is enhanced by a factor of 3 in the x2 orbits (Ginard et al. 2015). This result is consistent with NGC 5195 being a postburst system, with forbidden neon lines present in the IR spectrum (Boulade et al. 1996) and the optical lines (Hα absorption and strong [N ii] emission; Greenawalt et al. 1998) in the nucleus, but with little ionized gas in the outskirts. Deep IFU imaging of the optical lines (Hα and [N ii]) will be able to confirm this scenario.

We assume that the HCN emission is emerging from the same region as the CN emission and then estimated the line ratio by converting the CARMA flux to the appropriate brightness temperature for the NRO 45 m. We followed the same procedure for CS(2–1). We find line ratios of HCN(1–0)/CS(2–1) = 1.2, CN(1–0)/HCN(1–0) = 2.1 and CN(1–0)/CS(2–1) = 3.3. Comparing different transitions of different species is not ideal, because one is tracing a mixture of excitation and chemistry. In spite of this, it seems safe to say that CN is more abundant than HCN and CS in the central kpc of NGC 5195.

Using RADEX (van der Tak et al. 2007),¹⁶ and assuming a gas kinetic temperature of 65 K (from the dust temperature derived by Smith 1982), an HCN/CS ratio of about unity can be obtained with volume densities of about 10⁴ cm⁻³ and assuming similar HCN and CS abundances. This may point to a very low fraction of dense gas (i.e., gas with number density n ≳ 10⁵ cm⁻³) that is in agreement with the emission of CN and CS, as well as with the high ratio CO(1-0)/HCN(1-0) of 85.3.¹⁷

Assuming that CN is tracing PDRs in NGC 5195, the UV fields created by the low star formation in the galaxy (see Section 4) could dissociate and fragment the molecular clouds, creating larger envelopes in which most of the CN would form. The CN(1–0) line is brighter than that of HCN. The radiative transfer of CN is not straightforward, but if we assume collisional excitation, the line ratio indicates higher CN abundances that that of HCN. Since CN can be a photo-dissociation product of HCN, this may be the result of UV irradiation of fragmented dense clumps. However, one would then also expect to see a strong HCO⁺ line. There are some star-forming regions with X(CN) > X(HCN) > X(HCO⁺) in (for example) the Orion ridge (Blake et al. 1986), but instead this ratio may emerge from the region around an AGN. Lepp & Dalgarno (1996) suggested that X(CN) > X(HCN) in X-ray irradiated regions. It is possible that (at least a significant fraction of) the high density tracer emission is emerging from the nucleus instead of being scattered in a poststarburst region.

Apart from the species previously mentioned, we marginally detect CH₃CN $({6}_{k}-{5}_{k})$ (located between 200 and 350 km s⁻¹ at the red side of ¹³CO in Figure 8). Its integrated flux, of 1.3 Jy km s⁻¹ is, however, below a 3σ detection. This molecule is also a tracer of dense gas (${n}_{\mathrm{crit}}$[CH₃CN $({6}_{k}-{5}_{k})]\sim {10}^{5}$ cm⁻³), similar to those of CN(1–0) and CS(2–1). However, CH₃CN is easily dissociated by UV fields (Aladro et al. 2013), and thus it typically resides in the cores of the molecular clouds. Based on that, the ratio between CN and CH₃CN would be a rough indicator of the amount of dense gas residing in the envelopes versus the inner parts. We obtain a tentative ratio of CN/CH₃CN = 9.0. Interestingly, the circumnuclear regions of the Seyfert galaxies NGC 1097 and NGC 1068 (Aladro et al. 2013; Martí et al. 2015) show luminous CH₃CN emission. High-resolution observations will reveal if NGC 5195 also has a circumnuclear region of dense gas irradiated by an AGN. Such a structure may have a complex chemistry with suppressed HCO⁺ while CN abundances can be high and also those of CH₃CN in warm, shielded regions.

In order to classify the stellar population of NGC 5195, we used the Penalized Pixel Fitting software (ppxf; Cappellari & Emsellem 2004)¹⁸ applied to the nuclear spectrum¹⁹ of NGC 5195 from the SDSS DR12 (Alam et al. 2015). The $3^{\prime\prime}$ spectral fiber is ideal for this study, as spectra of NGC 5195 that are more extended than the nuclear region tend to be contaminated by the spiral arm of M51a (see: Figure 13.64 of Moustakas et al. 2010). Integral field spectroscopy would be able to differentiate between NGC 5195 and M51a, to gain a clearer picture of the integrated stellar population properties of NGC 5195, but is beyond the scope of this paper.

We used the stellar templates from the miles library (Vazdekis et al. 2010), normalized to a uniform (Salpeter 1955) initial mass function (IMF), using models representing 50 stellar ages from 0.06–17 Gyr, and four metallicities (−0.71 < Z < 0.22), for a total of 200 models. We included regularization within ppxf (regul = 0.004) to allow us to understand the range of possible stellar populations that would match the nuclear spectrum. Figure 10 shows the model fit to the central spectrum of NGC 5195, and Figure 11 shows the probability distribution function of the ppxf model fitting, summed over the metallicity axis. The nucleus of NGC 5195 clearly shows two stellar populations: 80% of nuclear stars are old (≳10 Gyr) and 20% are of intermediate age (≈1 Gyr), consistent with A-stars.

Zoom In Zoom Out Reset image size

As Figure 11 shows, there is some uncertainty to the age of the intermediate-aged stellar population, mostly based on degeneracies of metallicity, but also likely because the SDSS fiber only subtends a small portion (144 pc) of the nucleus. This does put a significant constraint on the interaction that took place between NGC 5195 and M51a, suggesting when enhanced star formation was present in NGC 5195. Given that star formation is known to increase even at large radii during gravitational encounters (Scudder et al. 2012; Moreno et al. 2015), it is likely that this stellar population represents the enhanced star formation that was taking place throughout the encounter (which was suggested to take place ≈1/2 Gyr ago; Salo & Laurikainen 2000; Dobbs et al. 2010). The spectral stellar population fit also agrees within errors with the photometrically derived stellar populations (Mentuch Cooper et al. 2012), and might provide another constraint to model the complex history of interaction that has occurred in the M51 system.

The SFR of NGC 5195 was calculated by Lanz et al. (2013) using magphys (da Cunha et al. 2008) to model the far-UV to far-IR SED, measuring it to be 0.142 ${M}_{\odot }$ yr⁻¹. To estimate the contribution of the M51a to the Lanz et al. (2013) SFR, we defined a region based on the 8 μm PAH emission from the foreground arm and measured the 70 and 8 μm photometry excluding the region. We find that 8% of the PACS 70 μm flux and 18% of the IRAC 8 μm flux are contained in the excluded region. While this region may also contain some emission from NGC 5195, the strong correlation of the luminosities in these bands to SFR enable us to estimate that 10%–20% of the Lanz et al. (2013) SFR may be due to M51a, revising the integrated SFR down to 0.116 M${}_{\odot }$ yr⁻¹.

Using the revised SFR, and dividing by the moment0 area from Table 2, we derive an SFR rate surface density ${{\rm{\Sigma }}}_{\mathrm{SFR}}$ ≈ 0.105 ${M}_{\odot }$ yr⁻¹ kpc⁻² (renormalized to our chosen distance to NGC 5195 and use of a Salpeter initial mass function; Salpeter 1955).

To determine if the molecular gas is inefficient globally, we investigate where NGC 5195 fits on the Schmidt–Kennicutt (S–K) star-formation surface density—gas surface density relation (Schmidt 1959; Kennicutt 1998; Kennicutt & Evans 2012):

$$\begin{eqnarray}&&\displaystyle \frac{{{\rm{\Sigma }}}_{\mathrm{SFR}}}{{M}_{\odot }\,{\mathrm{yr}}^{-1}\,{\mathrm{kpc}}^{-2}}=2.5\times {10}^{-4}{\left(\displaystyle \frac{{{\rm{\Sigma }}}_{{{\rm{H}}}_{2}}}{{M}_{\odot }{\mathrm{pc}}^{-2}}\right)}^{1.4}.\end{eqnarray} \tag{ 4 }$$

If we use the average molecular surface density calculated from our ¹²CO observations, $\langle {{\rm{\Sigma }}}_{{{\rm{H}}}_{2}}\rangle =87$ ±  2 ${M}_{\odot }$ pc⁻², we predict an average ${{\rm{\Sigma }}}_{\mathrm{SFR},\mathrm{predicted}}$ ≈ 0.13 ${M}_{\odot }$ yr⁻¹ kpc⁻², in good agreement with the measured ${{\rm{\Sigma }}}_{\mathrm{SFR}}$. We, therefore, conclude that the global molecular gas in NGC 5195 is efficiently forming stars. Two caveats to this analysis are (1) that the Lanz et al. (2013) star formation could have been contaminated by the spiral arm in M51a (seen in PAH emission in Figure 1) and (2) that the intermediate-aged stellar population could be providing a nonzero contribution to the far-IR luminosity. We further investigate the SFR to determine if either of these caveats play a significant role.

We also evaluated the SFR indicated by the centimeter- and millimeter-wave radio continuum emission using the following calibration to the radio-SFR relation from Murphy et al. (2011, 2013):

$$\begin{eqnarray}&&\left(\displaystyle \frac{{\mathrm{SFR}}_{\nu }}{{\text{}}{M}_{\odot }\,{\mathrm{yr}}^{-1}}\right)={10}^{-27}\\ &&\times \left[2.18{\left(\displaystyle \frac{{T}_{\mathrm{ex}}}{{10}^{4}{\rm{K}}}\right)}^{0.45}{\left(\displaystyle \frac{\nu }{\mathrm{GHz}}\right)}^{-0.1}\,+\,15.1{\left(\displaystyle \frac{\nu }{\mathrm{GHz}}\right)}^{{\alpha }_{\mathrm{NT}}}\right]\\ &&\quad \times \left(\displaystyle \frac{{L}_{\nu }}{\mathrm{erg}\,{{\rm{s}}}^{-1}\,{\mathrm{Hz}}^{-1}}\right),\end{eqnarray} \tag{ 5 }$$

with excitation temperature ${T}_{\mathrm{ex}}$ = 10⁴ K and the nonthermal spectral index ${\alpha }_{\mathrm{NT}}$ = −0.8. From the sensitive interferometric imaging provided by the SINGS project (Kennicutt et al. 2003; Braun et al. 2007), NGC 5195 has an integrated 1.4 GHz flux density of ∼17 mJy within the ¹²CO-detected region. This corresponds to a 1.4 GHz SFR of ∼0.14 ${M}_{\odot }$ yr⁻¹, slightly higher than the revised Lanz et al. (2013) integrated SFR. We, therefore, conclude that the centimeter-wave radio emission associated with NGC 5195 is predominantly produced by SF, rather than radio-loud AGN activity. This suggests that, if the nucleus of NGC 5195 is indeed currently active (as the molecular lines seem to suggest), it may be operating in the so-called radio-quiet or "high-excitation" mode characteristic of efficiently accreting massive black holes (Heckman & Best 2014), and consistent with the weak X-rays reported by Schlegel et al. (2016). Alternatively, AGN signatures at optical and X-ray wavelengths previously reported by Moustakas et al. (2010) and Terashima & Wilson (2004), respectively, may in fact be related to stellar processes or shocks occurring in the vicinity of the NGC 5195 nucleus (Lisenfeld & Völk 2010). If such a scenario in which the massive black hole in the center of NGC 5195 is in a quiescent state is indeed true, it is not clear what mechanism is responsible for halting AGN fueling given the abundant availability of cold gas. Additional detailed studies of the gas kinematics in the ambient environment of the NGC 5195 nucleus will be needed to disentangle these possibilities in the future.

At the level of the integrated flux density of ∼1 mJy of the compact 106 GHz emission, the SFR predicted by Equation (5) is ∼0.07 ${M}_{\odot }$ yr⁻¹. This is a lower limit, as it is likely that CARMA resolved out the diffuse, lower-level emission (seen in the 1.4 GHz map in Braun et al. 2007). Even with this caveat, the predicted SFR based on the millimeter continuum emission is similarly consistent with a pure SF origin.

In order to correctly differentiate star formation taking place in the molecular gas in NGC 5195, we use the reduced 70 μm image (see: Section 2.1) from Herschel PACS (Poglitsch et al. 2010). The map was then converted to a ${{\rm{\Sigma }}}_{\mathrm{SFR}}$ map using Equation (21) from Calzetti et al. (2010). We were able to directly compare this map to the ${{\rm{\Sigma }}}_{{{\rm{H}}}_{2}}$ map, which we derived from the ¹²CO(1–0) using Equation (3), after converting the unclipped integrated intensity map using the Kelvin per Jansky factor for the CARMA map (listed in Table 2). We then re-gridded the 70 μm map to match the ¹²CO map using the idl task hastrom.²⁰ We used the clipped moment maps (Figure 7) to define the comparison region (which contains "real" emission far above the noise of the map, with a total area of 3.3 kpc²).

We first calculated the total SFR by summing the ${{\rm{\Sigma }}}_{\mathrm{SFR}}$ corresponding to ¹²CO emission, SFR${}_{\mathrm{tot}}$ = 0.072 M${}_{\odot }$ yr⁻¹, that is a factor of 1.6 smaller than the revised SFR in Section 4.2, and consistent with the 106 GHz free–free estimate for the SFR. This is possibly due to the smaller area subtended by the detected ¹²CO emission. We then calculated the resolved depletion time for NGC 5195 by dividing the matched ${{\rm{\Sigma }}}_{{{\rm{H}}}_{2}}$ by ${{\rm{\Sigma }}}_{\mathrm{SFR}}$, shown in Figure 12. The average depletion timescale (taken only from unmasked pixels with positive values) is 3.08 Gyr (with a range between 1 and 6 Gyr).

Zoom In Zoom Out Reset image size

When we include helium, this depletion time increases to 4.31 Gyr. This depletion timescale is a factor of $\approx 2$ larger than that found in normal star-forming galaxies (Bigiel et al. 2011; Saintonge et al. 2011; Leroy et al. 2013). On the other hand, this depletion time is more consistent with the molecular gas depletion time found for ATLAS${}^{3{\rm{D}}}$ early-type galaxies (Davis et al. 2014). Given that NGC 5195 is an early-type galaxy, the depletion timescale consistent with early-type galaxies is fitting.

Overall, despite having taken part in the interactions with M51a a few hundred Myr ago, the gas in NGC 5195 appears to have resettled and returned to forming stars at normal efficiency (for an early-type galaxy). This provides an update to the observations of Kohno et al. (2002), who suggested that the molecular disk was gravitationally stable against collapse, suppressing star formation in this system. Our data suggest that NGC 5195 has returned to forming stars normally for its type (and its molecular gas content), which might include shear forces capable of reducing the star-formation efficiency by ∼2 (Martig et al. 2013; Davis et al. 2014), consistent with other early-type galaxies.

The level of star formation we currently observe in NGC 5195 is consistent with a star-formation history that includes an enhancement in star formation corresponding to a near approach and subsequent interaction with M51a (possibly including a super-efficient burst of star formation) that has declined for the past 1 Gyr, followed by a resettling of the molecular disk and the re-establishment of efficient, normal-mode star formation consistent with its early-type classification.