Seyfert galaxy NGC 7172. From top to bottom: Panel a: HST optical to near-IR composite RGB image of NGC 7172 retrieved from ESASky4.3.0. It combines observations from the Advanced Camera for Surveys (ACS) at 6060 and 8140 Å, and the Wide Field Camera 3 (WFC3) at 1.6 µm. The orientation of the image is north up, east to the left. The FoV is 2 × 2 . The rectangle shows approximately the ALMA FoV, while the dashed lines represent the opening angle of the southern (brighter) side of the optical ionization cone. Image credit: ESA/Hubble & NASA, D. J. Rosario, A. Barth. Panel b: ALMA CO(3−2) intensity map (moment 0) produced with a 3σ clipping using the merged configuration observations. We trimmed the original FoV to approximately 17.6 × 3.3 . Both the image and the contours are shown on a linear scale. The color bar shows the intensity scale in Jy km s −1 per beam units. The first contour is at 0.05 Jy km s −1 per beam. The ALMA beam (not shown) is 0.08 × 0.07 (14 pc × 13 pc) at PA beam = 90.4 • . Panel c: ALMA CO(3−2) mean velocity map (moment 1) produced with a 3σ clipping using the merged configuration observations. The contours are the isovelocities. Panel d: ALMA CO(3−2) mean velocity dispersion map (moment 2) produced with a 3σ clipping using the merged configuration observations. For the last two panels the scale bar units are km s −1 .

Seyfert galaxy NGC 7172. From top to bottom: Panel a: HST optical to near-IR composite RGB image of NGC 7172 retrieved from ESASky4.3.0. It combines observations from the Advanced Camera for Surveys (ACS) at 6060 and 8140 Å, and the Wide Field Camera 3 (WFC3) at 1.6 µm. The orientation of the image is north up, east to the left. The FoV is 2 × 2 . The rectangle shows approximately the ALMA FoV, while the dashed lines represent the opening angle of the southern (brighter) side of the optical ionization cone. Image credit: ESA/Hubble & NASA, D. J. Rosario, A. Barth. Panel b: ALMA CO(3−2) intensity map (moment 0) produced with a 3σ clipping using the merged configuration observations. We trimmed the original FoV to approximately 17.6 × 3.3 . Both the image and the contours are shown on a linear scale. The color bar shows the intensity scale in Jy km s −1 per beam units. The first contour is at 0.05 Jy km s −1 per beam. The ALMA beam (not shown) is 0.08 × 0.07 (14 pc × 13 pc) at PA beam = 90.4 • . Panel c: ALMA CO(3−2) mean velocity map (moment 1) produced with a 3σ clipping using the merged configuration observations. The contours are the isovelocities. Panel d: ALMA CO(3−2) mean velocity dispersion map (moment 2) produced with a 3σ clipping using the merged configuration observations. For the last two panels the scale bar units are km s −1 .

Source publication
Article
Full-text available
We present new ALMA observations of the CO(3−2) transition and associated 854 μm continuum at 0.06 − 0.3″ resolution, together with new VLT/SINFONI observations of NGC 7172. This is a luminous (bolometric luminosity of ≃10 ⁴⁴ erg s ⁻¹ ) Seyfert galaxy that belongs to the Galaxy Activity, Torus, and Outflow Survey (GATOS). The ALMA CO(3−2) observati...

Contexts in source publication

Context 1
... is also a broader dust lane extending east-west across the galaxy that appears to be warped at the edges ( Sharples et al. 1984). Figure 1 (top panel) shows a red-greenblue (RGB) optical to near-IR Hubble Space Telescope (HST) image of NGC 7172 retrieved from ESASky ( Baines et al. 2017;Giordano et al. 2018), where this morphology is clearly seen. The distribution of the prominent dust lanes to the north of the nucleus indicates this is the near side of the galaxy. ...
Context 2
... Thomas et al. (2017) detected a prominent two-sided ionization cone in NGC 7172 using optical integral field unit (IFU) spectroscopy. It has a projected opening angle of approximately 120 • . The southern part of the cone is brighter, which is consistent with the position of the main dust lane and with north being the near side of the galaxy (see Fig. 1, top panel). The size on this side of the cone is approximately 5 , which is equivalent to a projected physical size of 900 pc. Using spectroscopic observations with X-shooter on the Very Large Telescope (VLT), Davies et al. (2020) derived a velocity for the ionized gas outflow of ∼400 km s −1 and an outflow rate of 0.005 M yr −1 . In ...
Context 3
... sensitivities in the line data cubes were derived by selecting areas free from emission in all channels. We summarize in Table 1 all the relevant parameters of the ALMA observations. We created CO(3−2) moment maps from the merged configuration data set using the GILDAS moment task with a 3σ clipping. We show the 0th, 1st, and 2nd moment maps in Fig. 1, while Fig. 2 show the 854 µm continuum maps of the central region observed with the compact and merged configurations. Additionally, in Sect. 4 we discuss the moment maps from the compact ...
Context 4
... ALMA CO(3−2) and 854 µm emission Figure 1 (second panel from the top) shows the ALMA CO(3−2) intensity map (moment 0) derived from the merged configuration data set. It maximizes the best angular resolution with the best sensitivity of our observations. ...
Context 5
... of 6−8 or 1.1−1.4 kpc, containing bright clumps. Beyond the ring, the CO(3−2) emission becomes more diffuse and covers the ALMA FoV along the east-west direction. The presence of this ring was already suspected from previous ALMA CO(2−1) images at lower angular resolution (see Alonso-Herrero et al. 2020). For a scale reference, the top panel of Fig. 1 displays the approximate FoV of the ALMA CO(3−2) image plotted as a red rectangle. The CO(3−2) emission clearly traces the prominent dust lane that crosses the (circum)nuclear region of the galaxy. In the innermost 4 , the CO(3−2) emission delineates the extinction in the ring derived from the VLT/SINFONI near-IR continuum and the ...
Context 6
... map of the CO(3−2) mean velocity (see Fig. 1, third panel from the top) from the merged configuration shows a rotation pattern. There are also clear distorsions in the velocity field, which might be in part due to the presence of an ILR in NGC 7172. However the clear departure from rotation in central region along the minor axis of the galaxy are indicative of the presence of ...
Context 7
... the presence of an ILR in NGC 7172. However the clear departure from rotation in central region along the minor axis of the galaxy are indicative of the presence of radial motions. In Sect. 4, we perform a detailed modeling of the CO(3−2) kinematics and quantify the velocity and sign of these radial motions. The mean velocity dispersion map (see Fig. 1, bottom panel) shows a complex structure with several regions with velocity dispersions above 100 km s −1 , H 2 ...
Context 8
... (p−v) diagrams along the kinematic major and minor axes. Before we started the fits, we trimmed the original compact configuration data cube to the FoV with detected emission, ∼20.9 × 4.6 . Figure 6 shows the ALMA CO(3−2) observed moment maps of NGC 7172. The intensity, velocity, and velocity dispersion maps are similar to those in Fig. 1, considering the difference in angular resolution. The CO(3−2) mean velocity field shows a general rotation pattern. However, some important distorsions are readily seen in the molecular gas ring, which are accompanied by the presence of high velocity dispersion values, especially along the east-west direction in the disk of the ...
Context 9
... the approximate inner 2 , the minor axis p−v diagram shows redshifted motions to the south of the AGN and blueshifted to the north (see Fig. 8, bottom panel). Since the north is the near side of the galaxy (see the top panel of Fig. 1), this suggests the presence of an outflow in the cold molecular gas ring, if these motions are taking place in the plane of the galaxy. We thus run another 3D BAROLO model including a radial velocity (v rad ) component. We fixed the inclination to the average value from the previous run and the derived systemic velocity. We also fixed ...
Context 10
... Fig. 9 (top and bottom panels, respectively). It is immediately clear from the comparison with Fig. 7, that the 3D BAROLO model with the fixed rotation curve and the radial velocity component produces a better fit to the observations. The effect of including the latter is appreciated in the CO(3−2) p−v diagram along the kinematic minor axis (see Fig. 10). However, even when including radial velocities, the model does not reach the observed velocity of nearly +150 km s −1 to the south, close to the AGN position. The maximum fitted radial velocities are 80−100 km s −1 (see the bottom panel of Fig. 11), progressively decreasing to 20−30 km s −1 out to a radial distance of 4.5 . This ...
Context 11
... including the latter is appreciated in the CO(3−2) p−v diagram along the kinematic minor axis (see Fig. 10). However, even when including radial velocities, the model does not reach the observed velocity of nearly +150 km s −1 to the south, close to the AGN position. The maximum fitted radial velocities are 80−100 km s −1 (see the bottom panel of Fig. 11), progressively decreasing to 20−30 km s −1 out to a radial distance of 4.5 . This deceleration might be due to the AGN wind encountering large amounts of molecular gas. The bright peaks of H 2 , CO(3−2), and 854 µm emission in the ring (see Sect. 3) might be caused by this impact. Within the gas ring, the gas velocity dispersion takes ...
Context 12
... wind encountering large amounts of molecular gas. The bright peaks of H 2 , CO(3−2), and 854 µm emission in the ring (see Sect. 3) might be caused by this impact. Within the gas ring, the gas velocity dispersion takes values of approximately σ gas 10 km s −1 , except in the region with the largest radial velocity values (see the middle panel of Fig. 11) where the gas dispersion is slightly lower. At r > 4.5 , the velocity dispersion goes down to the minimum value imposed in our fit (σ gas = 5 km s −1 ...
Context 13
... the northeast. They appear to be coincident with those regions where the innermost part of AGN wind, traced by the VLT/SINFONI [Si vi] emission (see left panels of Fig. 4), impacts the inner part of the cold molecular gas ring. Even though at the radial distances of these regions we fit the largest values of the radial velocities (bottom panel of Fig. 11), the residuals mean that this azimuthally averaged model cannot explain completely the observed kinematics. In fact, these distinct velocity residuals also define kinematically the expanding molecular ring with a diameter of ∼6 or approximately 1.1 kpc. The residual mean velocities are not only confined to these regions, but there are ...
Context 14
... left panels of Fig. 12 show the observed 854 µm continuum of the central 0.4 × 0.4 (approximately 72 pc × 72 pc) for the observations with the highest angular resolution (extended configuration, top panel) and the intermediate resolution (merged configuration, bottom panel). The corresponding physical resolutions are approximately 11 pc and 13 pc, ...
Context 15
... to the synchrotron emission. Briefly, we started by fitting the extended configuration 854 µm continuum map with a point source with the observation beam size and orientation, and an elliptical component. The estimated unresolved flux is 520 µJy. We then subtracted the point-source fit from the observations. The point-source subtracted image (Fig. 12, top-right panel) shows no emission residuals at the AGN position, that is, all the observed emission at this angular resolution is unresolved. Next, we generated a point source image corresponding to the beam of the merged configuration data set, scaled it to the point source flux from the previous step, and subtracted it from the ...
Context 16
... corresponding to the beam of the merged configuration data set, scaled it to the point source flux from the previous step, and subtracted it from the observations. The 854 µm point source subtracted image shows significant extended emission at the 3 to 5 × σ levels, whereas at the AGN position there is no residual emission (bottomright panel of Fig. 12). The measured fluxes are 877 µJy for the point + extended emission and 364 µJy after the point source subtraction. The extended emission therefore accounts for approximately 42% of the observed flux within r = 32 pc and Notes. We used different brightness temperature ratios for the torus and the other regions, namely, T B CO ( can be ...
Context 17
... value estimated from the cold dust emission. It is also within the range of torus masses inferred in other Seyfert galaxies ( García-Burillo et al. 2021). We measured an H 2 column density at the AGN position of log (N H2 ) = 22.34 mol cm −2 , which agrees well the observed values in other GATOS Seyfert galaxies with similar AGN luminosities (see Fig. 19, right, from García-Burillo et al. ...
Context 18
... Sect. 4.2, we showed that the CO(3−2) minor axis p−v diagram (Fig. 10) goes across significantly strong emission from the ring. Under the commonly assumed hypothesis that the bulk of molecular gas shares a coplanar geometry, the v rad term is the only component of gas motions that has a nonzero projection along the disk minor axis. If the molecular gas is following circular orbits, for the known ...
Context 19
... absorption feature. They measured an approximate blueshifted velocity of 50 km s −1 of the OH feature, which the authors considered just at the limit of their outflow detection criterion. This value of the OH velocity is similar to the typical value of the CO(3−2) radial velocity fit with 3D BAROLO in the molecular gas ring (see bottom panel of Fig. 11). In the particular case of NGC 7172, the detection of the OH outflow was probably aided by the fact that the outflow is observed close to our line of ...
Context 20
... quantity explaining the different gas phases of outflows in active galaxies ( Fluetsch et al. 2019;Davies et al. 2020). On the other hand, our results indicate that NGC 7172 follows well the trend between L AGN and molecular gas mass outflow rate found by Lutz et al. (2020) for a compilation of local infrared bright galaxies, gray points in Fig. 13. We note however that other works, such as Ramos and Lamperti et al. (2022), do not find tight correlations between these quantities for samples of type 2 quasars (QSO2s) and local ULIRGs classified as AGNs, red triangles and orange squares, respectively. In simulations of AGN winds in disk galaxies, the outflows are predicted to ...
Context 21
... vi] emission (Sect. 3 and Figs. 2 and 4) add support to this scenario. At this particular location of the NGC 7172 ring (also on the other side, to the northeast of the AGN), it is likely that the impact with material in the ring caused a strong deceleration of the AGN wind and subsequently in the outflow velocity of the molecular gas, as seen in Fig. 11 (bottom panel). Indeed, in Davies et al. (2020) sample, NGC 7172 is among the Seyfert galaxies with the lowest velocities of the ionized gas, thus providing further evidence to the above ...
Context 22
... r = 200 pc scale does not encompass the entire ring at all. Increasing the size of the circumnuclear region to include the ring gives a concentration index of log Σ r=50 pc H2 /Σ r=720 pc H2 = −0.6. The comparison of the two molecular gas concentration indices of NGC 7172 with the sample of low-luminosity AGNs and GATOS Seyfert galaxies (Fig. 18 of García-Burillo et al. 2021) indicates that NGC 7172 follows well the observed trends for the two physical scales of the torus and its immediate surroundings. That is, there is a redistribution of the molecular gas in the torus and nuclear regions of NGC 7172 when compared to other less luminous Seyfert galaxies and AGNs. For this ...
Context 23
... finally explore the fate of the circumnuclear molecular gas in NGC 7172 since its mass outflow rate appears to be at the high end of those observed in other AGNs of similar luminosity (Fig. 13). On the other hand, the velocity of the molecular outflow of NGC 7172 is modest when compared with the galaxy circular velocity. In Appendix B we modeled the escape velocity in NGC 7172 using the 3D BAROLO rotation curve. At a radial distance of 900 pc, just outside of the outflow region, the escape velocity is 850 km s −1 , whereas at ...
Context 24
... outflow of NGC 7172 is modest when compared with the galaxy circular velocity. In Appendix B we modeled the escape velocity in NGC 7172 using the 3D BAROLO rotation curve. At a radial distance of 900 pc, just outside of the outflow region, the escape velocity is 850 km s −1 , whereas at smaller radial distances this velocity is even higher (see Fig. B.1). Given that the molecular outflow is taking place mostly in the disk of the galaxy and its velocity is low with respect to the expected escape velocity, we can therefore conclude that in this galaxy most of the molecular gas will not leave the galaxy, at least during this period of AGN ...
Context 25
... appendix describes a bar-ring model, with the nuclear bar being the source of noncircular motions in the center, as an alternative explanation for the observed CO(3-2) kinematics of the central regions of NGC 7172. The central gas distribution mapped with ALMA has an extent of 3.2 kpc or projected radius of 1.6 kpc (c.f., Fig. 1). Its morphology reveals a central ring, with the characteristic enhancement in the northeast and southwest parts of the ring, as shown in Fig. 2. This indicates that the ring may be a pseudo-ring, made by the winding of the two spiral arms, elongated along the bar, and usually observed as the dusty leading edges of the bar in spiral ...
Context 26
... ring is ∼300 pc, typical of an ILR. This central molecular feature appears very elongated. It is likely however that the ring and barred structure are intrinsically elongated, considering that the inclination of the galaxy is not completely edge-on, but of the order of ∼60 • , as noted by Hyperleda and the observed large-scale optical morphology (Fig. 1). We then assumed this value of the inclination in the following modelization. We used N-body simulations with gas, star formation, and feedback, as described in Combes (2008). The Poisson equation is solved with a 3D Particle-Mesh code based on FFT, and is fully self-consistent with a live dark halo. The gas dissipation is represented ...
Context 27
... several components, including a stellar bulge and disk, a dark matter halo, and a gas disk, with the parameters listed in Table A.1. The shape of these components are Miyamoto-Nagai disks and Plummer spheres, as described in Chilingarian et al. (2010). The computed circular velocity and angular frequencies, Ω, Ω ± κ/2, Ω − κ/4, are plotted in Fig. ...

Similar publications

Preprint
Full-text available
We present statistics of $z\sim 6-9$ galaxy outflows indicated by spatially-extended gas emission and broad lines. With a total of 61 spectroscopically confirmed galaxies at $z\sim 6-9$ in the JWST CEERS, GLASS, and ERO data, we find five galaxies with [O{\sc iii}]+H$\beta$ ionized gas emission significantly extended beyond the kpc-scale stellar co...
Article
Full-text available
We present the first systematic study of the molecular gas and star formation efficiency in a sample of ten narrow-line Seyfert 1 galaxies selected to have X-ray Ultra Fast Outflows and, therefore, to potentially show AGN feedback effects. CO observations were obtained with the IRAM 30m telescope in six galaxies and from the literature for four gal...
Article
Full-text available
Context. The overwhelming majority of diagnostic tools for galactic activity are focused mainly on the classes of active galaxies. Passive or dormant galaxies are often excluded from these diagnostics, which usually employ emission-line features (e.g., forbidden emission lines). Thus, most of them focus on specific types of activity or only on one...
Article
Full-text available
The physical quantity that directly controls the feedback of active galactic nuclei (AGNs) in elliptical galaxies remains to be determined. The discovery of molecular gas around the AGNs suggests that the gas is fueling the AGNs. Therefore, we analyze Atacama Large Millimeter/submillimeter Array data for the CO line ( J = 1–0, 2–1, 3–2) emission an...
Article
Full-text available
A large fraction of the accreting supermassive black hole population is shrouded by copious amounts of gas and dust, particularly in the distant ( z ≳ 1) universe. While much of the obscuration is attributed to a parsec-scale torus, there is a known contribution from the larger-scale host galaxy. Using JWST/NIRCam imaging from the COSMOS-Web survey...

Citations

... The molecular gas in galaxies is the fuel for star formation (SF) and is af fected by acti v e galactic nuclei (AGNs) feedback processes (F abian 2012 ; K ennicutt & Ev ans 2012 ; Cattaneo 2019 ), such as accretion (Combes et al. 2013 ;Izumi, Kawakatu & Kohno 2016 ;Farrah et al. 2022b ) and/or outflows (the review by Veilleux et al. 2020 , and more recently Garc ía-Bernete et al. 2021 ; Ramos Almeida et al. 2022 ;Alonso Herrero et al. 2023 ). Most of the molecular gas resides in giant molecular clouds (GMCs) that are the birthplaces of stars and, for this reason, influence the o v erall chemical and dynamical evolution of a galaxy (Hoyle 1953 ;Tan 2000 ;Ballesteros-Paredes et al. 2020 ). ...
Article
Full-text available
We present a new physically-motivated model for estimating the molecular line emission in active galaxies. The model takes into account (i) the internal density structure of giant molecular clouds (GMCs), (ii) the heating associated both to stars and to the active galactic nuclei (AGN), respectively producing photodissociation regions (PDRs) and X-ray dominated regions (XDRs) within the GMCs, and (iii) the mass distribution of GMCs within the galaxy volume. The model needs, as input parameters, the radial profiles of molecular mass, far-UV flux and X-ray flux for a given galaxy, and it has two free parameters: the CO-to-H2 conversion factor αCO, and the X-ray attenuation column density NH. We test this model on a sample of 24 local (z ≤ 0.06) AGN-host galaxies, simulating their carbon monoxide spectral line energy distribution (CO SLED). We compare the results with the available observations and calculate, for each galaxy, the best (αCO, NH) with a Markov chain Monte Carlo algorithm, finding values consistent with those present in the literature. We find a median αCO = 4.8 M⊙ (K km s−1 pc2)−1 for our sample. In all the modelled galaxies, we find the XDR component of the CO SLED to dominate the CO luminosity from Jupp ≥ 4. We conclude that, once a detailed distribution of molecular gas density is taken into account, PDR emission at mid-/high-J becomes negligible with respect to XDR.
Article
Full-text available
We present new optical GTC/MEGARA seeing-limited ($0.9 integral-field observations of NGC 5506, together with ALMA observations of the CO($3-2$) transition at a $0.2 pc) resolution. NGC 5506 is a luminous (bolometric luminosity of $ $ erg s$^ $) nearby (26 Mpc) Seyfert galaxy, part of the Galaxy Activity, Torus, and Outflow Survey (GATOS). We modelled the CO($3-2$) kinematics with 3D Barolo revealing a rotating and outflowing cold gas ring within the central 1.2 kpc. We derived an integrated cold molecular gas mass outflow rate for the ring of $ odot $ yr$^ $. We fitted the optical emission lines with a maximum of two Gaussian components to separate rotation from non-circular motions. We detected high OIII 5007$ projected velocities (up to $ 1000$ km $) at the active galactic nucleus (AGN) position, decreasing with radius to an average $ 330$ km $ around $ pc. We also modelled the OIII gas kinematics with a non-parametric method, estimating the ionisation parameter and electron density in every spaxel, from which we derived an ionised mass outflow rate of $0.076$ M$_ odot $ yr$^ $ within the central 1.2 kpc. Regions of high CO($3-2$) velocity dispersion, extending to projected distances of $ 350$ pc from the AGN, appear to be the result from the interaction of the AGN wind with molecular gas in the galaxy's disc. Additionally, we find the ionised outflow to spatially correlate with radio and soft X-ray emission in the central kiloparsec. We conclude that the effects of AGN feedback in NGC 5506 manifest as a large-scale ionised wind interacting with the molecular disc, resulting in outflows extending to radial distances of 610 pc.
Article
Full-text available
We use JWST/MIRI MRS spectroscopy of a sample of six local obscured type 1.9/2 active galactic nuclei (AGN) to compare their nuclear mid-IR absorption bands with the level of nuclear obscuration traced by X-rays. This study is the first to use subarcsecond angular resolution data of local obscured AGN to investigate the nuclear mid-IR absorption bands with a wide wavelength coverage (4.9-28.1 mu m). All the nuclei show the 9.7\,mu m silicate band in absorption. We compare the strength of the 9.7 and 18\,mu m silicate features with torus model predictions. The observed silicate features are generally well explained by clumpy and smooth torus models. We report the detection of the 6\,mu m dirty water ice band (i.e., a mix of water and other molecules such as CO and CO$_2$) at subarcsecond scales (sim 0.26 at 6\,mu m; inner sim 50\,pc) in a sample of local AGN with different levels of nuclear obscuration in the range log N$_ H X-Ray $(cm$^ We find good correlation between the 6\,mu m water ice optical depths and H X-Ray $. This result indicates that the water ice absorption might be a reliable tracer of the nuclear intrinsic obscuration in AGN. The weak water ice absorption in less obscured AGN (log N$_H^ X-ray $ (cm$^ $)lesssim 23.0 cm$^ $) might be related to the hotter dust temperature ($>$T$_ sub H_2O expected to be reached in the outer layers of the torus due to their more inhomogeneous medium. Our results suggest it might be necessary to include the molecular content, such as H$_2$O, aliphatic hydrocarbons (CH-), and more complex polycyclic aromatic hydrocarbon (PAH) molecules, in torus models to better constrain key parameters such as the torus covering factor (i.e., nuclear obscuration).
Article
Full-text available
We analyze NIRSpec observations of the CO rovibrational $v$=1--0 band at sim 4.67 around the dust-embedded southern active galactic nucleus (AGN) of NGC 3256 ($d$=40\,Mpc; IR $\,$L_ We classify the CO $v$=1--0 spectra into three categories based on the behavior of P- and R-branches of the band: (a) both branches in absorption toward the nucleus; (b) P-R asymmetry (P-branch in emission and R-branch in absorption) along the disk of the galaxy; and (c) both branches in emission in the outflow region above and below the disk. In this paper, we focus on the outflow. The CO $v$=1--0 emission can be explained by the vibrational excitation of CO in the molecular outflow by the bright mid-IR sim 4.7 continuum from the AGN up to $r$sim 250\,pc. We model the ratios between the P($J$+2) and R($J$) transitions of the band to derive the physical properties (column density, kinetic temperature, and CO-to-H$_2$ conversion factor, $ CO $) of the outflowing gas. We find that the 12CO $v$=1--0 emission is optically thick for $J$<4, while the 13CO $v$=1--0 emission remains optically thin. From the P(2) R(0) ratio, we identify a temperature gradient in the outflow from $>$40\,K in the central 100\,pc to $<$15\,K at 250\,pc, sampling the cooling of the molecular gas in the outflow. We used three methods to derive $ CO $ in eight 100\,pc (0 apertures in the outflow by fitting the P($J$+2) R($J$) ratios with nonlocal thermodynamic equilibrium (NLTE) models. We obtain low median $ CO $ factors (0.40--0.61)$ CO H_2 $\,$M_ (K $ in the outflow regions. This implies that outflow rates and energetics might be overestimated if a 1.3--2 times larger ultraluminous infrared galaxy (ULIRG) like $ CO $ is assumed. The reduced $ CO $ can be explained if the outflowing molecular clouds are not virialized. We also report the first extragalactic detection of a broad (sigma =0.0091 spectral feature at 4.645 associated with aliphatic deuterium on polycyclic aromatic hydrocarbons (D$_n$-PAHs).