A&A 615, A33 (2018) https://doi.org/10.1051/0004-6361/201732178 Astronomy c ESO 2018 & Astrophysics

Diffuse interstellar bands λ5780 and λ5797 in the Antennae Galaxy as seen by MUSE A. Monreal-Ibero1,2, P. M. Weilbacher3, and M. Wendt3,4

1 Instituto de Astrofísica de Canarias (IAC), 38205 La Laguna, Tenerife, Spain e-mail: [email protected] 2 Universidad de La Laguna, Dpto. Astrofísica, 38206 La Laguna, Tenerife, Spain 3 Leibniz-Institut für Astrophysik Potsdam (AIP), An der Sternwarte 16, 14482 Potsdam, Germany 4 Institut für Physik und Astronomie, Universität Potsdam, Karl-Liebknecht-Str. 24/25, 14476 Golm, Germany Received 26 October 2017 / Accepted 30 January 2018

ABSTRACT

Context. Diffuse interstellar bands (DIBs) are faint spectral absorption features of unknown origin. Research on DIBs beyond the Local Group is very limited and will surely blossom in the era of the Extremely Large Telescopes. However, we can already start paving the way. One possibility that needs to be explored is the use of high-sensitivity integral field spectrographs. Aims. Our goals are twofold. First, we aim to derive reliable mapping of at least one DIB in a galaxy outside the Local Group. Second, we want to explore the relation between DIBs and other properties of the interstellar medium (ISM) in the galaxy. Methods. We use Multi Unit Spectroscopic Explorer (MUSE) data for the Antennae Galaxy, the closest major galaxy merger. High signal-to-noise spectra were created by co-adding the signal of many spatial elements with the Voronoi binning technique. The emission of the underlying stellar population was modelled and substracted with the STARLIGHT spectral synthesis code. Flux and equivalent width of the features of interest were measured by means of fitting to Gaussian functions. Results. To our knowledge, we have derived the first maps for the DIBs at λ5780 and λ5797 in galaxies outside the Local Group. 2 The strongest of the two DIBs (at λ5780) was detected in an area of 0.60, corresponding to a linear scale of 25 kpc . This region was sampled using >200 out of 1200 independent lines of sight.∼ The DIB λ5797 was detected in >100 independent∼ lines of sight. Both DIBs are associated with∼ a region of high emission in the H i 21 cm line, implying a connection between atomic gas and DIBs, as the correlations in the Milky Way also suggest. Conversely, there is mild spatial association between the two DIBs and the molecular gas, in agreement with results for our Galaxy that indicate a lack of correlation between DIBs and molecular gas. The overall structures for the DIB strength distribution and extinction are comparable. Within the system, the λ5780 DIB clearly correlates with the extinction, and both DIBs follow the relationship between equivalent width and reddening when data for several galaxies are considered. This relationship is tighter when comparing only with galaxies with metallicities close to solar. Unidentified infrared emission bands (UIBs, likely caused by polycyclic aromatic hydrocarbons PAHs) and the λ5780 and λ5797 DIBs show similar but not identical spatial distributions. We attribute the differences to extinction effects without necessarily implying a radically different nature of the respective carriers. Conclusions. The results illustrate the enormous potential of integral field spectrographs for extragalactic DIB research. Key words. dust, extinction – ISM: lines and bands – galaxies: ISM – galaxies: individual: Antennae Galaxy – galaxies: interactions – ISM: structure

1. Introduction remain unidentified. Hydrocarbon chains (Maier et al. 2004; Motylewski et al. 2000; Krełowski et al. 2011), polycyclic Stellar spectra display a plethora of weak absorption features aromatic hydrocarbons (PAHs; Leger & D’Hendecourt 1985; ff called di use interstellar bands (DIBs; see Herbig 1995; Sarre van der Zwet & Allamandola 1985; Salama et al. 1996), and 2006, for reviews). These were first identified by Heger(1922) fullerenes (Kroto 1988; Sassara et al. 2001; Iglesias-Groth 2007) and their interstellar origin was established a few years later are among the possible candidates. In general, carbon seems to (Merrill 1936). Today, we know more than 400 optical DIBs (e.g. be involved and in that sense, DIBs have been proposed as the Jenniskens & Desert 1994; Hobbs et al. 2009). Additionally, 30 ∼ largest reservoir of organic matter in the Universe (Snow 2014). DIBs have been detected in the near-infrared (e.g. Joblin et al. The number of viable candidates can be reduced by study- 1990; Geballe et al. 2011; Cox et al. 2014; Hamano et al. 2016; ing the DIBs’ intrinsic characteristics (e.g. band profiles) and Elyajouri et al. 2017) while there is no firm detection in the near- the relationships between different DIBs and the interstellar UV yet (Bhatt & Cami 2015). medium (ISM), in general. Thus, it was found that most of the Identification of the carriers (i.e. molecules causing the strong DIBs are correlated (e.g. Friedman et al. 2011; Kos et al. absorptions) has proven to be an arduous task. With the excep- 2013; Puspitarini et al. 2013), though this is not always the case + tion of C60, recently confirmed as the carrier of the DIBs at for weaker DIBs, which can even show a negative correlation λ9577 and λ9632 (Campbell et al. 2015; Walker et al. 2015, but (Baron et al. 2015). However, with the possible exception of the also see Galazutdinov & Krełowski 2017 for objections to λ6196.0–λ6613.6 pair, none of the derived correlations can be this identification), those for the vast majority of DIBs considered perfect. Likewise, they also correlate – to different

Article published by EDP Sciences A33, page 1 of 12 A&A proofs: manuscript no. aanda

ent extents depending on the DIB – with extinction (e.g. Herbig Schweizer et al. 2008). The system has a total infrared luminos- 1993), neutral hydrogen, interstellar Na I D and Ca H&K lines ity of log(Lir/L ) = 10.86 (Sanders et al. 2003) approaching that (e.g. Lan et al. 2015; Friedman et al. 2011), and other molecules of a luminous infrared galaxy, and is subject of heavy and vari- in the ISM (e.g. CH, C2, CN, Weselak et al. 2004; Baron et al. able attenuation (Brandl et al. 2005). It therefore presents a good 2015). In contrast, they are poorly correlated with the amount of target for the search of extragalactic DIBs. H2 (e.g. Lan et al. 2015; Friedman et al. 2011). Besides, differ- The paper is structured as follows: Sect. 2 summarises the ent DIBs react in different ways to the UV radiation field (e.g. data collection and reduction; Sect. 3 contains a description of Krelowski et al. 1992; Cami et al. 1997; Cox & Spaans 2006; our methods to extract the information for the DIBs and other Vos et al. 2011; Cordiner et al. 2013; Elyajouri et al. 2017) and used observables. Results are presented and discussed in Sect. 4. seem to show small scale spatial variations (Wendt etA&A al. 2017;615, A33Our (2018) main conclusions are summarised in Sect. 5. Monreal-Ibero et al. 2015a; Cordiner et al. 2013; van Loon et al. 2009) and different levels of clumpiness (Kos 2017). extents depending on the DIB – with extinction (e.g. Herbig log10F(5200−5600) An important piece of information is the behaviour of DIBs +5.25 1993), neutral hydrogen, interstellar Na I D and Ca H & K lines outside of the Milky Way (see Cordiner 2014, for an excellent re- 40 (e.g. Lan et al. 2015; Friedman et al. 2011), and other molecules 2 kpc view of the subject). Galaxies other than ours span a large diver- in the ISM (e.g. CH, C , CN, Weselak et al. 2004; Baron et al. 20 sity in terms of stellar,2 gas and dust content, mass, morphology, 2015). In contrast, they are poorly correlated with the amount of +4.75 evolutionary status, environmental conditions, amongst others. H (e.g. Lan et al. 2015; Friedman et al. 2011). Besides, differ- 0 + 2This entails a diversity of ISM properties, and thus an invalu- ent DIBs react in different ways to the UV radiation field (e.g. able opportunity to test the existence and properties of DIBs un- −20 Krełowski et al. 1992; Cami et al. 1997; Cox & Spaans 2006; der gas physical (e.g. radiation field, temperature, density) and Vos et al. 2011; Cordiner et al. 2013; Elyajouri et al. 2017) and +4.25 chemical (e.g. metallicity, relative abundances, molecular con- −40 seem to show small scale spatial variations (Wendt et al. 2017; (arcsec) tent) conditions and dust properties not found in the Milky Way. Monreal-Ibero et al. 2015a; Cordiner et al. 2013; van Loon et al. ∆δ −60 However, since they are faint features, works aiming to study 2009) and different levels of clumpiness (Kos 2017). + extragalactic DIBs are still scarce. +3.75 An important piece of information is the behaviour of DIBs −80 outsideThe of firstthe Milky DIB detected Way (see outside Cordiner the 2014 Milky, for Way anwas excellent that at λ −100 review4428, of in the the subject). Magellanic Galaxies Clouds other (Hutchings than ours 1966; span Blades a large & Madore 1979). Nowadays, there are certainly many more DIB diversity in terms of stellar, gas and dust content, mass, mor- −120 +3.25 phology,detections evolutionary within these status, galaxies environmental (Ehrenfreund conditions, et al. 2002;amongst Cox et al. 2006, 2007; Welty et al. 2006; Bailey et al. 2015; van 80 60 40 20 0 −20 −40 −60 others. This entails a diversity of ISM properties, and thus an ∆α (arcsec) invaluableLoon et al. opportunity 2013), reaching to test up the to several existence tens and of lines properties of sight for the DIBs at λ5780 and λ5797. In both galaxies, DIBs seem of DIBs under gas physical (e.g. radiation field, temperature, Fig. 1. Line-free continuum as derived from STARLIGHT showing the weaker with respect to N(H i) and E(B V) than in our Galaxy. Fig. 1. Line-free continuum as derived from STARLIGHT showing the density) and chemical (e.g. metallicity,− relative abundances, mapped area with the tessellated pattern used to extract the high signal- The lower metallicity of the systems and, perhaps, the increased to-noisemapped ratio area spectra with the (i.e. tessellated the tiles). Thepattern reconstructed used to extract white-light the high image signal- molecular content) conditions and dust properties not found in to-noise ratio spectra (i.e. the tiles). The reconstructed white-light image theradiation Milky Way. field However, have been since proposed they as are explanations faint features, (Cox works et al. is overplotted as reference with contours in logarithmic stretching of 2006; Welty et al. 2006). Additionally, these same DIBs were 0.5is dex overplotted steps. The as origin reference of coordinates with contours here in and logarithmic through stretchingthe paper was of 0.5 aiming to study extragalactic DIBs are still scarce. dex steps. The origin of coordinates here and through the paper was put alsoThefirstDIBdetectedoutsidetheMilkyWaywasthatat measured in a few lines of sight for the Andromedaλ4428, galaxy put at the position of the northern nucleus. North is up, east is to the left.at the position of the northern nucleus. North is up, east is to the left. in(M the 31, Magellanic Cordiner et Clouds al. 2011) (Hutchings and in the 1966 Triangulum; Blades galaxy & Madore (M 33, 1979Cordiner). Nowadays, et al. there2008). are Outside certainly of many the Local more DIB Group, detections DIB de- tections are scant. They have been found in relatively nearby within these galaxies (Ehrenfreund et al. 2002; Cox et al. 2006, mergers as well as being the closest (distance = 22 3 Mpc, galaxies that show high extinction (e.g Ritchey & Wallerstein 2007; Welty et al. 2006; Bailey et al. 2015; van Loon et al. 2013), Schweizer2. The data et al. 2008). The system has a total infrared± luminos- 2015; Heckman & Lehnert 2000), towards bright supernovae reaching up to several tens of lines of sight for the DIBs at λ5780 ity of log(Lir/L ) = 10.86 (Sanders et al. 2003) approaching that (D’Odorico et al. 1989; Thöne et al. 2009; Cox & Patat 2008; The dataset we use in this paper has already been presented in and λ5797. In both galaxies, DIBs seem weaker with respect to of a luminous infrared Galaxy, and is subject of heavy and vari- Welty et al. 2014; Sollerman et al. 2005), and damped Lyman-α great detail in Weilbacher et al. (2017, hereafter Paper I). Here, N(H i) and E(B V) than in our Galaxy. The lower metallicity of able attenuation (Brandl et al. 2005). It therefore presents a good systems (e.g.− Junkkarinen et al. 2004; Srianand et al. 2013). we give a short summary of its properties and the processing the systems and, perhaps, the increased radiation field have been target for the search of extragalactic DIBs. As demonstrated in some of the examples mentioned above, involved. proposed as explanations (Cox et al. 2006; Welty et al. 2006). Ad- The paper is structured as follows: Sect.2 summarises the highly multiplexed spectrographs offer the possibility of collect- The central merger of the NGC 4038/39 system was ob- ditionally, these same DIBs were also measured in a few lines data collection and reduction; Sect.3 contains a description of ing in one shot information for many lines of sight. Integral served in April/May 2015 and February and May 2016 with of sight for the Andromeda Galaxy (M 31; Cordiner et al. 2011) our methods to extract the information for the DIBs and other field spectrographs (IFSs) constitute a special case of this kind MUSE on VLT UT4 (Bacon et al. 2012). The wide-field mode and in the Triangulum Galaxy (M 33; Cordiner et al. 2008). Out- used observables. Results are presented and discussed in Sect.4. of instrumentation where this information is recorded in an ex- in extended configuration was used, which yields a sampling side of the Local Group, DIB detections are scant. They have Our main conclusions are summarised in Sect.5. tended continuous field. In Monreal-Ibero et al. (2015b), we pro- of 000.2 over a field of about 10, with a wavelength range of been found in relatively nearby galaxies that show high extinc- 1 tionposed (e.g Ritchey the use of & high-sensitivity Wallerstein 2015 IFSs; Heckman as a tool & to Lehnert detect 2000 and map), 4600 to 9350 Å, sampled at about 1.25 Å pixel− . As explained 2. The data towardsDIBs outside bright supernovaethe Local Group. (D’Odorico The experiment, et al. 1989 conducted; Thöne et using al. in Paper I, we reduced the data using the public MUSE pipeline 2009the;Cox Multi & Unit Patat Spectroscopic2008;Welty et al.Explorer2014;Sollerman (MUSE) commissioninget al.2005), The(v. dataset 1.6; Weilbacher we use in et this al. paper 2014, has Weilbacher already been et al. presented in prep.), in us- anddata, damped led to Lyman- the firstα determinationsystems (e.g.of Junkkarinen a DIB radial et al. profile 2004 in; a greatingdetail standard in Weilbacher calibrations, et including al.(2018, bias hereafter subtraction, Paper flat-fielding I). Here, Srianandgalaxy et outside al. 2013 the). Local Group. The experiment demonstrated weand give slice a short tracing, summary wavelength of its calibration, properties twilight and the sky processing correction, theAs possibilities demonstrated of usingin some IFSs of inthe extragalactic examples mentioned DIB research, above, but involved.atmospheric refraction correction, and sky subtraction (using ex- the strength of the DIB had to be determined only as average ternalThe skycentral fields) merger as well of as the correction NGC 4038 to/39 barycentric system was velocities. ob- highly multiplexed spectrographs offer the possibility2 of col- lectingvalues in in one large shot areas information (i.e. between for 0.7 many and lines 3.8 kpc of sight.). Inte- servedAll exposures in April/May of the 2015 central and field February were then and combined, May 2016 using with the gral fieldBuilding spectrographs on top of that (IFSs) experiment, constitute we a havespecial initiated case of a search this MUSElinear on full VLT width UT4 at (halfBacon maximum et al. 2012 (FWHM)). The wide-field weighting mode scheme kindfor of DIBs instrumentation in the galaxies where observed this information within the is MUSE recorded Guaran- in inbuilt extended into the configuration pipeline, to optimise was used, the which final seeing. yields We a sampling minimised anteed extended Time continuous Observations. field. Here, In Monreal-Ibero we present the et results al.(2015b for), the ofcontributions 000.2 over a field from of exposures about 1 of0, withlower a sky wavelength transparency range by of re- Antennae Galaxy (NGC 4038/NGC 4039, Arp 244). This sys- setting their FWHM keywords to high values.1 Most of the expo- we proposed the use of high-sensitivity IFSs as a tool to de- 4600 to 9350 Å, sampled at about 1.25 Å pixel− . As explained tecttem and is map one ofDIBs the outside most spectacular the Local Group. examples The of experiment, gas-rich ma- in suresPaper were I, we 1350 reduced s long, the databut we using replaced the public one smallMUSE section pipeline that jor mergers as well as being the closest (distance=22 3 Mpc, was saturated at Hα by two 100 s exposures. The final output conducted using the Multi Unit Spectroscopic Explorer (MUSE)± (v. 1.6; Weilbacher et al. 2014, and in prep.), using standard cal- commissioning data, led to the first determination of a DIB radial ibrations, including bias subtraction, flat-fielding and slice trac- Article number, page 2 of 12 profile in a galaxy outside the Local Group. The experiment ing, wavelength calibration, twilight sky correction, atmospheric demonstrated the possibilities of using IFSs in extragalactic refraction correction, and sky subtraction (using external sky DIB research, but the strength of the DIB had to be deter- fields) as well as correction to barycentric velocities. All expo- mined only as average values in large areas (i.e. between 0.7 sures of the central field were then combined, using the linear and 3.8 kpc2). full width at half maximum (FWHM) weighting scheme built Building on top of that experiment, we have initiated a search into the pipeline, to optimise the final seeing. We minimised con- for DIBs in the galaxies observed within the MUSE Guaran- tributions from exposures of lower sky transparency by resetting teed Time Observations. Here, we present the results for the their FWHM keywords to high values. Most of the exposures Antennae Galaxy (NGC 4038/NGC 4039, Arp 244). This sys- were 1350 s long, but we replaced one small section that was sat- tem is one of the most spectacular examples of gas-rich major urated at Hα by two 100 s exposures. The final output datacube

A33, page 2 of 12 A. Monreal-Ibero et al.: Diffuse interstellar bands λ5780 and λ5797 in the Antennae Galaxy as seen by MUSE A. Monreal-Ibero et al.: Diffuse interstellar bands λ5780 and λ5797 in the Antennae Galaxy as seen by MUSE

Spectrum 0850 Spectrum 0303

1.3 1.2 1.2 1.1 1.1 1.0 1.0 0.9 0.9 0.8 0.8 5000 5500 6000 6500 5000 5500 6000 6500 λ(Å) λ(Å) 1.15 1.15 1.15 1.2 1.2 1.2 1.10 1.10 1.10 1.1 1.1 1.1 1.05 1.05 1.05 1.00 1.00 1.00 1.0 1.0 1.0 0.95 0.95 0.95 5780 5797 5780 5797 0.9 0.9 λ λ 0.9 0.90 0.90 λ λ 0.90 4700 4800 4900 5000 5100 5680 5742 5805 5868 5930 6510 6540 6570 6600 6630 4700 4800 4900 5000 5100 5680 5742 5805 5868 5930 6510 6540 6570 6600 6630 λ(Å) λ(Å) λ(Å) λ(Å) λ(Å) λ(Å)

Spectrum 0076 Spectrum 0096 2.0 1.5 1.5

1.0 1.0

0.5 0.5 0.0 0.0 5000 5500 6000 6500 5000 5500 6000 6500 λ(Å) λ(Å) 1.4 1.4 1.4 1.4 1.4 1.4 1.2 1.2 1.2 1.2 1.2 1.2 1.0 1.0 1.0 1.0 1.0 1.0 0.8 0.8 0.8 0.8 0.8 0.8 0.6 0.6 0.6 0.6 0.6 0.6 0.4 0.4 0.4

5780 5797 0.4 0.4 5780 5797 0.4 0.2 0.2 λ λ 0.2 λ λ 4700 4800 4900 5000 5100 5680 5742 5805 5868 5930 6510 6540 6570 6600 6630 4700 4800 4900 5000 5100 5680 5742 5805 5868 5930 6510 6540 6570 6600 6630 λ(Å) λ(Å) λ(Å) λ(Å) λ(Å) λ(Å)

Fig. 2. Examples modelling stellar populations. Each panel displays the complete modelled spectral range in the upper row and three zooms inFig. key 2. spectralExamples regions modelling in the stellarlower populations. row. The observed Each panel spectrum displays is in the black complete and the modelled modelled spectral stellar range spectral in the energy upper distribution row and three in light zooms blue. in Maskedkey spectral spectral regions ranges in the are lower marked row. using The observedbands with spectrum green stripes is in black in diagonal. and the modelledThe positions stellar of spectral the DIBs energy at λ5780 distribution and λ5797 in light are marked blue. Masked in the centralspectral zoom ranges of are each marked panel with using red bands ticks. withUpper green row stripes: two red in diagonal. spectra without The positions DIB detection. of the DIBsLower at rowλ5780: two and blueλ5797 spectra are without marked DIB in the detection. central Reportedzoom of each wavelengths panel with here red and ticks. throughtUpper the row: paperTwo are inred rest-frame. spectra without DIB detection. Lower row: Two blue spectra without DIB detection. Reported wavelengths here and throught the paper are in rest-frame. has 950 973 4045 voxels1, with an effective seeing of about 6800 Å (rest-frame). We used a set of simple stellar populations × × 0datacube.6 FWHM hasat 950 7000 Å.973 Due4045 to the voxels irregular1, with layout an e offfective the obser- see- ofcase, three we di modelledfferent metallicities the spectral (Z range= 0.004, from 0.02, 4570 and Å 0.05) to 6800 from Å 00 × × vations,ing of about some 000 of.6 the FWHM spatial at regions 7000 Å. in Due the cubeto the remain irregular empty. layout A Bruzual(rest-frame). & Charlot We used(2003 a set) which of simple are based stellar on populations the Padova of 2000 three reconstructedof the observations, continuum some image of the of spatial the mapped regions area in the used cube in this re- evolutionarydifferent metallicities tracks (Girardi (Z=0.004, et al. 0.02, 2000 and) and 0.05) assumed from Bruzual a Salpeter & workmain empty.is displayed A reconstructed as contours continuum in Fig.1. Please image see of Paper the mapped I for initialCharlot mass (2003) function which between are based 0.1 on and the 100 PadovaM . 2000For each evolution- metal- otherarea used information in this work about is the displayed data. as contours in Fig. 1. Please licity,ary tracks there (Girardi was a total et al. of 2000) 15 spectra and assumed with ages a ranging Salpeter from initial 1 see Paper I for other information about the data. mass function between 0.1 and 100 M . For each metallicity, Myr to 13 Gyr. Critical spectral ranges corresponding to ionised gasthere emission was a total lines, of Wolf-Rayet 15 spectra star with emission, ages ranging possible from ISM 1 Myr ab- 3. Extracting information sorptionto 13 Gyr. features, Critical and spectral sky substractionranges corresponding residuals to were ionised masked gas 3. Extracting information emission lines, Wolf-Rayet star emission, possible ISM absorp- out for the fit. Finally, we assumed a common extinction for all 3.1. Tile definition and modelling of the underlying stellar tion features, and sky substraction residuals were masked out for 3.1. Tile definition and modelling of the underlying stellar the base spectra, and modelled it as a uniform dust screen with population emission the fit. Finally, we assumed a common extinction for all the base population emission the extinction law by Cardelli et al.(1989). The cube was tessellated by grouping and co-adding the spectra spectra, and modelled it as a uniform dust screen with the extinc- The cube was tessellated by grouping and co-adding the spectra All the main outputs for each tile were then reformatted in using the Voronoi binning technique (Cappellari & Copin 2003) tion law by Cardelli et al. (1989). using the Voronoi binning technique (Cappellari & Copin 2003) order to create four files with the total, modelled stellar emis- to produce data of signal to noise ratio S/N 250 in the mostly All the main outputs for each tile were then reformatted in or- to produce data of S/N 250 in the mostly∼ emission-line free sion, residuals, and ratio of total to stellar emission. Addition- emission-line free wavelength∼ range 5600–6500 Å. Then, the der to create four files with the total, modelled stellar emission, wavelength range 5600. . . 6500 Å. Then, the stellar spectral en- ally, the information associated to auxiliary properties such as stellar spectral energy distribution in each tile was modelled and residuals, and ratio of total to stellar emission. Additionally,2 the ergy distribution in each tile was modelled and subtracted. This stellar extinction in the V band, the reduced χ and the abso- subtracted. This is necessary to properly derive the fluxes for the information associated to auxiliary properties such as stellar ex- is necessary to properly derive the fluxes for the Balmer emis- lute deviation of the fit (in %) were reformatted2 and saved as 2D Balmer emission lines, and crucial to identify the faint ISM fea- tinction in the V band, the reduced χ and the absolute deviation sion lines, and crucial to identify the faint ISM features in ab- FITS images suitable for manipulation with standard astronom- tures in absorption. To do so, we used the STARLIGHT2 spectral of the fit (in %) were reformatted and saved as 2D FITS images sorption. To do so, we used the STARLIGHT2 spectral synthesis ical software. Use both map and image to refer to this kind of synthesis code (Cid Fernandes et al. 2005, 2009) that reproduces suitable for manipulation with standard astronomical software. code (Cid Fernandes et al. 2005, 2009) that reproduces a given file. a given observed spectrum by selecting a linear combination of use both map and2 image to refer to this kind of file. observed spectrum by selecting a linear combination of a sub-set The reduced χ was relatively uniform all over the covered a sub-set of N spectral components from a pre-defined set of areaThe with reduced a meanχ2 (wasstandard relatively deviation) uniformχ all2 of over 2.23 the ( covered0.43). of N spectral∗ components from a pre-defined set of base spec- base∗ spectra. Our aim was not to in detail determine the charac- ± χ2 ± tra. Our aim was not to in detail determine the characteristics of Thearea withsame a applies mean ( tostandard the absolute deviation) deviationof of2.23( the0.43). fit, which The teristics of the stellar population but to recover the information ± ± the stellar population but to recover the information for the ISM wassame of applies0.4%. to the To absolute have an deviation idea of ofthe the tessellation fit, which patternwas of for the ISM component, both in emission and absorption. In our ∼ component, both in emission and absorption. In our particular for0.4%. the To mapped have an area, ideaof a theline-free tessellation continuum pattern is for presented the mapped in particular case, we modelled the spectral range from 4570 Å to area,∼Fig.1. a Some line-free representative continuum examplesis presented of in the Fig. modelled 1. Some spectra repre- 1 In this paper, we only used the linearly sampled version of the cube. sentativeare found examples in Figs.2 of and the3. modelled Each panel spectra displays are found the complete in Figs. 2 12 Inhttp://starlight.ufsc.br/index.php?section=1 this paper, we only used the linearly sampled version of the cube. andmodelled 3. Each spectral panel displays range in the the complete upper row modelled and three spectral zooms range in 2 http://www.astro.ufsc.br/en/starlight/ key spectral regions in the lower row. The positions of the DIBs Article number, page 3 of 12 A33, page 3 of 12 A&A proofs: manuscript no. aanda A&A 615, A33 (2018)

Spectrum 0157 Spectrum 0203 1.4 1.8 1.6 1.2 1.4 1.2 1.0 1.0 0.8 0.8 0.6 0.4 5000 5500 6000 6500 5000 5500 6000 6500 λ(Å) λ(Å) 1.3 1.3 1.3 1.6 1.6 1.6 1.2 1.2 1.2 1.4 1.4 1.4 1.1 1.1 1.1 1.2 1.2 1.2 1.0 1.0 1.0 1.0 1.0 1.0

0.9 0.9 5780 5797 0.9 0.8 0.8 5780 5797 0.8 λ λ λ λ 4700 4800 4900 5000 5100 5680 5742 5805 5868 5930 6510 6540 6570 6600 6630 4700 4800 4900 5000 5100 5680 5742 5805 5868 5930 6510 6540 6570 6600 6630 λ(Å) λ(Å) λ(Å) λ(Å) λ(Å) λ(Å)

Spectrum 0222 Spectrum 0265 2.0 2.0

1.5 1.5

1.0 1.0

0.5 0.5 5000 5500 6000 6500 5000 5500 6000 6500 λ(Å) λ(Å) 1.8 1.8 1.8 1.6 1.6 1.6 1.6 1.6 1.6 1.4 1.4 1.4 1.4 1.4 1.4 1.2 1.2 1.2 1.2 1.2 1.2 1.0 1.0 1.0 1.0 1.0 1.0 0.8 0.8 0.8 5780 5797 0.8 0.8 5780 5797 0.8 0.6 0.6 λ λ 0.6 λ λ 4700 4800 4900 5000 5100 5680 5742 5805 5868 5930 6510 6540 6570 6600 6630 4700 4800 4900 5000 5100 5680 5742 5805 5868 5930 6510 6540 6570 6600 6630 λ(Å) λ(Å) λ(Å) λ(Å) λ(Å) λ(Å)

Fig. 3. Same as Fig.2 but for spectra with DIB detection. Upper row: two spectra with a clear residual only at the position of the DIB at λ5780. λ LowerFig. 3. rowSame: two as spectraFig. 2 but with for apparent spectrawith residuals DIB at detection. the positionUpper of both, row: Two the DIB spectra at λ5780 with aand clear that residual at λ5797. only at the position of the DIB at 5780. Lower row: Two spectra with apparent residuals at the position of both, the DIB at λ5780 and that at λ5797. at λ5780 and λ5797 are marked in the central enlargement of the resolution of the MUSE data, these features are a blend of eachin the panel. upper row and three zooms in key spectral regions in the severalto this set DIBs. of data. We Therefore, created a spectrum we decided using to postpone the DIB any compila- exper- lowerSpectra row. The in Fig. positions2 are examples of the DIBs with at noλ5780 apparent and detectionλ5797 are tioniment by with Jenniskens this feature & Desert for future(1994 studies.), which was degraded and re- ofmarked the DIBs in the at centralλ5780 enlargement and λ5797. These of each spectra panel. could be both: sampledWe reproduced to mimic the the characteristics DIBs at λ5780 of theand MUSEλ5797 data. as two We Gaus- then relativelySpectra red in (upper Fig. 2 row), are examplesor relatively with blue no (lower apparent row). detection Spectra measuredsian functions the relative fixing thecentral relative wavelengths wavelength there. between We allowed them. for At inof Fig. the3 DIBs are examples at λ5780 with and DIBλ5797. detections. These spectra Thesecould are all be spectra both: absorptionthe resolution features of the only, MUSE and data, assumed these the features same lineare a width blend for of withrelatively a red red continuum. (upper row), In the or relativelyupper row, blue only (lower the residual row). Spectra in the bothseveral features. DIBs. The We createdshape of a these spectrum features using is thedominated DIB compila- by the modellingin Fig. 3 are of examples the spectra with at λ DIB5780 detections. is apparent, These at first are sight. all spectral In the instrumentaltion by Jenniskens spectral & resolutionDesert (1994), of the which MUSE was data degraded and thus, and us- re- lowerwith a row red both continuum. residuals, In theat λ upper5780 and row,λ5797, only the are residual clearly seen. in the ingsampled Gaussian to mimic functions the characteristics is a reasonable of assumption. the MUSE data. We used We then the modelling of the spectra at λ5780 is apparent, at first sight. In the resultsmeasured for the the relative fit of the central Hβ emission wavelengths line there. in velocity We allowed and line for widthabsorption to establish features the only, initial and guess assumed for the the same central line wavelength width for 3.2.lower Spectral row both fitting residuals, at λ5780 and λ5797, are clearly seen. andboth width features. of the The absorption shape of features. these features Additionally, is dominated we included by the We fitted the different spectral features needed for our analy- ainstrumental one-degree spectral polynomial resolution to model of the the MUSE residuals data in and the thus, stellar us- sis3.2. using Spectral the IDL-based fitting routine mpfitexpr (Markwardt 2009). continuum.ing Gaussian functions is a reasonable assumption. We used the Both emission and absorption features were fitted to Gaussian resultsWe fordisplay the fit some of the examples Hβ emission of fits in line Fig. in4. velocity The upper and row line functions.We fitted the different spectral features needed for our analy- containswidth to the establish fits for thespectra initial in Fig. guess2, consistent for the central with no wavelength DIB de- sis usingThe fitting the IDL-based to the strong routine emissionmpfitexpr lines was(Markwardt done in the 2009). resid- tectionand width (see of Sect. the absorption 3.3). Still, features. one can Additionally, see a possible we detection included of a ualsBoth (“observed–modelled” emission and absorption) spectra. features The were procedure fitted tois relatively Gaussian theone-degree DIB at λ polynomial5780 in the spectrum to model 850the residuals just below in the the significance stellar con- standardfunctions. and does not pose a particular challenge for the goals level.tinuum. The lower row has the fits for the spectra in Fig.3. A feature λ of thisThe work. fitting Therefore, to the strong we emission refer to Monreal-Ibero lines was done et in al. the(2010 resid-) in absorptionWe display at the some position examples of the of DIB fits at in Fig.5797 4. is The clearly upper seen row in foruals the (”observed details about - modelled” it. ) spectra. The procedure is relatively allcontains four cases, the fits including for spectra spectra in Fig. 157 2, and consistent 203. Likewise, with no DIB the two de- standardThe fitting and does of the not faint pose absorption a particular features challenge is, in for turn, the goalsquite examplestection (see with Sect. strongest 3.3). Still, DIBs one display can see an additional a possible absorption detection ofat / λ 5810 Å. The wavelength is consistent with that of a blend challenging.of this work. This Therefore, was done we referin the to ratio Monreal-Ibero (“observed etmodelled” al. (2010)) the∼ DIB at λ5780 in the spectrum 850 just below the significance spectra.for the details In this about paper, it. we have focussed on the absorption fea- oflevel. two The fainter lower DIBs row (hasλλ5809.12,5809.96, the fits for the spectra Jenniskens in Fig. & 3. Desert A fea- turesThe at λ fitting5780 andof theλ5797. faint Nonetheless, absorption features we inspected is, in turn, the spec- quite 1994ture in). The absorption detection at of the this position additional of the absorption DIB at λ is5797 encouraging. is clearly trachallenging. for the presence This was of done other in the (stronger) ratio (”observed DIBs and/ modelled” found evi-) However,seen in all its four magnitude cases, including is comparable spectra to 157 the and standard 203. Likewise,deviation dencesspectra. of, In at this least, paper, the DIB we have at λ6284 focussed in many on the of them. absorption This DIB fea- ofthe the two residuals examples and with it will strongest not be DIBsdiscussed display further. an additional ab- requirestures at λ a5780 more and delicateλ5797. correction Nonetheless, of the we telluric inspected absorption the spectra and sorption at λ 5810 Å. The wavelength is consistent with that i λλ ∼ [Ofor] the6300,6364 presence of emission other (stronger) lines than DIBs the one and currently found evidences applied 3.3.of a blend Quantifying of two fainterbias and DIBs uncertainties (λλ5809.12,5809.96 Jenniskens & toof, this atleast, set of thedata. DIB Therefore, at λ6284 we in decided many ofto postpone them. This any DIB exper- re- Desert 1994). The detection of this additional absorption is en- imentquires with a more this delicate feature for correction future studies. of the telluric absorption and Wecouraging. simulated However, the bias its magnitudeand uncertainties is comparable of the to whole the standard global [O iWe]λλ6300,6364 reproduced emission the DIBs lines at λ5780 than the and oneλ5797 currently as two applied Gaus- proceduredeviation of by the means residuals of mock and it spectra will not realistically be discussed matching further. the sian functions fixing the relative wavelength between them. At STARLIGHT input spectra. The greatest source of uncertainty Article number, page 4 of 12 A33, page 4 of 12 A. Monreal-Ibero Monreal-Ibero et al.: Diffuseuse interstellar interstellar bands bands λλ57805780 and and λλ57975797 in in the the Antennae Antennae Galaxy Galaxy as as seen seen by by MUSE MUSE

Fit 0850 Fit 0303 Fit 0076 Fit 0096 1.00 1.00 1.00 1.00

0.95 0.95 0.95 0.95

0.90 0.90 0.90 0.90 Flux (a. u.) Flux (a. u.) Flux (a. u.) Flux (a. u.) 5780 5797 5809 5780 5797 5809 5780 5797 5809 5780 5797 5809 λ λ λ λ λ λ λ λ λ λ λ λ 0.85 EW(λ5780) = 51 mÅ 0.85 EW(λ5780) = 27 mÅ 0.85 EW(λ5780) = 0 mÅ 0.85 EW(λ5780) = 0 mÅ EW(λ5797) = 28 mÅ EW(λ5797) = 13 mÅ EW(λ5797) = 0 mÅ EW(λ5797) = 0 mÅ 0.80 0.80 0.80 0.80 5760 5780 5800 5820 5760 5780 5800 5820 5760 5780 5800 5820 5760 5780 5800 5820 Wavelength (Å) Wavelength (Å) Wavelength (Å) Wavelength (Å)

Fit 0157 Fit 0203 Fit 0222 Fit 0265 1.00 1.00 1.00 1.00

0.95 0.95 0.95 0.95

0.90 0.90 0.90 0.90 Flux (a. u.) Flux (a. u.) Flux (a. u.) Flux (a. u.) 5780 5797 5809 5780 5797 5809 5780 5797 5809 5780 5797 5809 λ λ λ λ λ λ λ λ λ λ λ λ 0.85 EW(λ5780) = 201 mÅ 0.85 EW(λ5780) = 304 mÅ 0.85 EW(λ5780) = 320 mÅ 0.85 EW(λ5780) = 274 mÅ EW(λ5797) = 67 mÅ EW(λ5797) = 86 mÅ EW(λ5797) = 143 mÅ EW(λ5797) = 124 mÅ 0.80 0.80 0.80 0.80 5760 5780 5800 5820 5760 5780 5800 5820 5760 5780 5800 5820 5760 5780 5800 5820 Wavelength (Å) Wavelength (Å) Wavelength (Å) Wavelength (Å)

Fig. 4. DIB fit for the spectra presented in Fig.2 2( (upper row) and Fig.3 3( (lower row). The ratio of observed to modelled spectra are plotted in black while the total fits are in blue. Red dashed lines mark the fitted continua.

account that the velocity of our spectra is a priori unknown by 500 these two DIBs in our mock spectra. To estimate the uncertain- 160 STARLIGHT.ties, the specific Finally, value weof this interpolated ratio is not the critical simulated so long spectra as the to 140 ) ) t 400 t the sampling of the MUSE spectra, added some random noise to

u u whole range of expected equivalent widths are covered. Next, the

p p 120 t t u u simulatespectra were the S reddened/N of the using STARLIGHT an attenuation input law spectra (Cardelli and scaled et al. O 300 O 100 ( (

) ) them so that the continuum around the DIBs is normalised to 0 7 1989) and assuming a similar relation between extinction and

8 9 80 7 7

5 200 5 one. The total number of mock spectra was 500. They were pro- 60 the λ5780 DIB as in our Galaxy. Following, we added random ( (

W W 40 cessed as described in Sects. 3.1 and 3.2, exactly as our MUSE E E offsets in wavelength of up to 1 Å, to take into account that 100 ± 20 spectra.the velocity of our spectra is a priori unknown by STARLIGHT. 0 0 The comparison between the input and measured equivalent 0 100 200 300 400 500 0 20 40 60 80 100120140160 Finally, we interpolated the simulated spectra to the sampling EW( 5780) (Input) EW( 5797) (Input) widthsof the MUSE for both spectra, DIBs in added the mock some spectra random is noise shown to in simulate Fig.5. The the Fig. 5. Results for mock data. Measured equivalent width in mÅ as figure shows that we are able to adequately recover the equiva- a function of the input in the mock data (left: EW(λ5780); right: S/N of the STARLIGHT input spectra and scaled them so that the Fig. 5. Results for mock data. Measured equivalent width in mÅ as λ EW(λ5797)). The continuous orange line shows the one-degree poly- lentcontinuum width for around the DIB theDIBs at 5780 is normalised with a standard to one. deviation The total for num- the a function of the input in the mock data (left: EW(λ5780); right: nomial fit while the red dashed line shows the one-to-one relationship. residuals of 40 mÅ. Still, at low equivalent width, our whole EW(λ5797)). The continuous orange line shows the one-degree poly- ber of mock∼ spectra was 500. They were processed as described Our adopted limit for DIB detection is marked with a dotted green line. procedurein Sect. 3.1 systematically and Sect. 3.2, overestimates exactly as our the MUSE equivalent spectra. width. In nomial fit while the red dashed line shows the one-to-one relationship. that sense, measured values .100 mÅ should be seen as upper Our adopted limit for DIB detection is marked with a dotted green line. ff λ of the DIB measurements is a possible contamination by a stellar limits.The Results comparison are slightly between di theerent input for and the measuredDIB at 5797 equivalent (stan- spectral feature at λ5782, with contributions from Fe i, Cr i, Cu i, dardwidths deviation for both of DIBs25 in mÅ). the mock This spectra feature is is shown intrinsically in Fig. fainter5. The i and thus more sensitive∼ to our procedure. The experiment shows and3.3. Mg Quantifying(Worthey bias et al. and 1994 uncertainties). The feature varies with both age figure shows that we are able to adequately recover the equiva- and metallicity of the underlying stellar population. We mea- thatlent width our measuring for the DIB procedure at λ5780 systematically with a standard overestimates deviation for the suredWe simulated the equivalent the bias width and of uncertainties this feature in of our the data, whole obtaining global equivalentresiduals of width40 for mÅ. this Still, DIB. at The low comparison equivalent width, of input our and whole out- typicalprocedure values by means of between of mock650 spectra and realistically800 mÅ. Taking matching this into the put equivalent∼ widths provides us with a correction factor (the ∼ ∼ procedure systematically overestimates the equivalent width. In account,STARLIGHT weinput useda spectra. linearcombination The greatest sourceof two ofspectra uncertainty from the of one-degreethat sense, measuredpolynomial values displayed.100 as mÅ an orange should continuous be seen as line up- basethe DIB library measurements (Bruzual & is Charlot a possible 2003 contamination) as starting point by a of stellar our inper Fig. limits.5) for Results our measurements. are slightly di Wefferent automatically for the DIB rejected at λ5797 any measured equivalent width smaller than the standard deviation simulations.spectral feature They at λ have5782, an with equivalent contributions width from of Fe700i, Cr mÅi, Cu fori, (standard deviation of 25 mÅ). This feature is intrinsically ∼ plus the zero-point for the one-degree polynomial fit between thisand Mg feature,i (Worthey and very et al. diff 1994).erent The ages feature ( 15 Myr varies and with2.5 both Gyr), age fainter and thus more sensitive∼ to our procedure. The experi- ∼ ∼ the input and output equivalent widths. This limit is shown by bothand metallicity with solar ofmetallicity. the underlying We note stellar that population.the specific We selection mea- ment shows that our measuring procedure systematically over- the green dotted line in Fig.5. Given the faintness of the spectral ofsured the the stellar equivalent population width is of not this critical feature as in long our data, as the obtaining equiva- estimates the equivalent width for this DIB. The comparison of features, all the remaining fits were inspected afterwards to avoid lenttypical width values of the of stellar between absorption650 and feature800 ismÅ. comparable Taking this to intothat input and output equivalent widths provides us with a correction ∼ ∼ the inclusion of spurious DIB detection in the analysis. The measuredaccount, we in used our spectra. a linear Thencombination we added of two two spectra Gaussian from func- the factor (the one-degree polynomial displayed as an orange con- remaining measurements are discussed in following sections. tionsbase libraryat the position (Bruzual of & the Charlot DIBs at 2003)λ5780 as and startingλ5797, point with of vary- our tinuous line in Fig. 5) for our measurements. We automatically Maps of the DIB distributions are a direct presentation of our ingsimulations. equivalent They widths have in anλ equivalent5780 from width 0 to of 500700 mÅ. mÅ The for ratio this rejected any measured equivalent width smaller than the stan- ∼ measurements without any correction, while the comparisons of betweenfeature, and the verytwo features different varies ages ( with15 theMyr environment and 2.5 Gyr), (see both e.g. dard deviation plus the zero-point for the one-degree polynomial ∼ ∼ the DIB strengths with other properties (i.e. reddening, measure- Voswith et solar al. 2011 metallicity.; Cox et We al. note2017 that) with the the specific DIB selectionat λ5797 of being the fit between the input and output equivalent widths. This limit is ments from literature) include a correction based on the one- typicallystellar population three to five is not times critical fainter. as Welong assumed as the equivalent a factor of width three shown by the green dotted line in Fig. 5. Given the faintness of degree polynomials derived above (i.e. orange line in Fig.5). betweenof the stellar these absorption two DIBs feature in our is mock comparable spectra. to To that estimate measured the the spectral features, all the remaining fits were inspected after- uncertainties,in our spectra. the Then specific we added value two of this Gaussian ratio is functions not critical at the so wards to avoid the inclusion of spurious DIB detection in the longposition as the of thewhole DIBs range at λ of5780 expected and λ5797, equivalent with varyingwidths are equiva- cov- 4.analysis. Results The and remaining discussion measurements are discussed in follow- ered. Next, the spectra were reddened using an attenuation law lent widths in λ5780 from 0 to 500 mÅ. The ratio between the 4.1.ing sections. Mapping Maps the DIBs of the in DIB the distributions system are a direct presenta- (twoCardelli features et al. varies 1989) with and the assuming environments a similar (see relation e.g. Vos between et al. tion of our measurements without any correction, while the com- extinction2011; Cox and et al. the 2017)λ5780 with DIB the as DIB in our at Galaxy.λ5797 being Following, typically we Theparisons maps of showing the DIB the strengths detections with of other DIBs properties at λ5780 (i.e. and redden-λ5797 addedthree to random five times offsets fainter. in wavelength We assumed of a up factor to of1 Å, three to take between into areing, presented measurements in Fig. from6. They literature) constitute include the main a correction outcome of based the ± Article number,A33, page 5 of 12 A&A proofs: manuscript no. aanda on the one-degree polynomials derived above (i.e. orange line in area, but also an important part of the northern spiral arm. This Fig. 5). A&A 615, A33suggests (2018) a relationshipA. Monreal-Ibero with the extinction, et al.: Diffuse as interstellar has been bandsfoundλ in5780 and λ5797 in the Antennae Galaxy as seen by MUSE the Milky Way and other nearby galaxies (see e.g. Welty et al. EW(λ5780) 2014, and references therein)VLA and− HI will 21 be cm explored in detail in man et al. 2011; Welty 2014). Still, the maximum of the DIB +400 +315 Sects. 4.440 and 4.5. absorption is found in a region with mild emission in CO, and 40 As reviewed in the introduction, studies mostly in the Milky both DIBs and CO are detected in the northern spiral arm. Way, but also in nearby galaxies, show that DIBs may or may 20 20 The lack of DIB detection at the location with strong CO not be related with other constituents of the ISM. In that sense, emission coupled with the detection in the places with mild CO +300 +235 0 + we can profit0 from the wealth of ancillary+ information about the emission in Arp 244 could be explained in the context of the skin Antennae Galaxy to explore if similar connections in this system effect, as in the Milky Way. This is interpreted as being that DIB −20 exist. This−20 is presented in the following sections. formation takes place most efficiently in the outer regions of the interstellar cloud. Additionally, extinction could play a role, as +200 +155 will be discussed in Sect. 4.4. −40 4.2. Neutral−40 atomic hydrogen and DIBs (arcsec) (arcsec) ∆δ ∆δ −60 The λ−605780 DIB correlates well with neutral hydrogen in our E(B−V) (gas) + Galaxy (e.g. Friedman et al. 2011)+ while for a given amout of +1.00 −80 +100 neutral−80 hydrogen, DIBs seems underabundant in the Magellanic +75 40 Clouds (Welty et al. 2006). Figure 7 shows the H i 21 cm line −100 map−100 with the VLA downloaded from the NED3. The original 20 . +0.75 map has a spatial resolution of 110040 (Hibbard et al. 2001). Here 0 + −120 +0 the map−120 has been resampled and binned matching our tessella- −5 80 60 40 20 0 −20 −40 −60 80 60 40 20 0 −20 −40 −60 tion, to ease the comparison with the DIB maps. Looking at this −20 ∆α (arcsec) map together with those presented∆α (arcsec) in Fig. 6 (and also Fig. 1), we +0.50 EW(λ5797) see that both the atomic hydrogen and DIB distributions do not −40 +200 Fig.necessarily 7. VLA matchH ii 21 cm that map of the((HibbardHibbard stars, et as al. al. seen 2001 2001) in) once once the optical. resampled Inter- and 40 binned.estingly, Units the λ are5780 arbitrary and λ and5797 the DIBs colour are scale detected is in linear in regions stretch. with The (arcsec) reconstructedreconstructed white-light white-light image image is is overplotted overplotted as as reference reference with with con- con- ∆δ −60 toursantours excess in in logarithmic logarithmic of neutral stretching stretching hydrogen of of 0.5 0.5 (part dex dex ofsteps. steps. the overlap area, north- + 20 ern spiral arm), supporting the connection between the carriers −80 +0.25 of these two DIBs and H i. Still, the atomic hydrogen extends + +150 Caltech Millimeter Array − CO 0 Antennaewell beyond Galaxy the region to explore with if detection similar connections of DIB absorption, in this system+0.40 as if −100 exist.the characteristic This40 is presented scale in lengths the following for these sections. were smaller. That is: −20 at a given location, existence of H i seems a necessary condi- −120 +0.00 +100 tion for20 the existence of DIB carriers, but not enough to ensure it 80 60 40 20 0 −20 −40 −60 −40 4.2.alone. Neutral Similar atomic results hydrogen have been and found DIBs in our Galaxy when −0.25 com- ∆α

(arcsec) (arcsec) Theparingλ5780 radial0 DIB profiles correlates of these well species with+ (Puspitarinineutral hydrogen et al. 2015).in our ∆δ −60 E(B−V) (stars) + GalaxyAn exception (e.g. Friedman is a narrow et al. tongue 2011) at while[ 10 for00, a given200] joining amout the of +1.00 −20 ∼ − − −80 +50 neutralnorthern hydrogen, nucleus with DIBs the seems overlap underabundant area. While absorption in the Magellanic in both 40 DIBs have been detected, the VLA H i 21 cm does noti show −0.90 any Clouds−40 (Welty et al. 2006). Figure7 shows the H 21 cm line −100 mapparticularly with the strong VLA emission downloaded there. from The the spatial NED resolution3. The original of the 20 VLA (arcsec) image, larger than the width of the tongue, may explain map has a spatial resolution of 1100.40 (Hibbard et al. 2001). Here +0.75 ∆δ −60 −120 +0 thethis map effect. has Our been DIB resampled detections and and binned the VLA matching map would our tessella- still be 0 + i + 80 60 40 20 0 −20 −40 −60 tion,compatible to ease with the comparison the need of with H for the having DIB maps. DIB absorptionLooking −1.55 at if this at this location−80 there were a localised concentration of H i with typ- −20 ∆α (arcsec) map together with those presented in Fig.6 (and also Fig.1), we ical size larger than the width of the tongue but smaller than the +0.50 see that−100 both the atomic hydrogen and DIB distributions do not −40 Fig. 6. Maps of the derived equivalent width in mÅ for DIBs at λ5780 necessarilyVLA resolution. match that of the stars, as seen in the optical. Inter- (top) and λ5797 (bottom). The reconstructed white-light image is over- estingly, the λ5780 and λ5797 DIBs are detected in regions −2.20 with (arcsec)

−120 ∆δ plotted as reference with contours in logarithmic stretching of 0.5 dex −60 an4.3. excess Molecular of80 neutral gas60 and hydrogen40 DIBs20 (part0 of the−20 overlap−40 area,−60 north- + steps. ern spiral arm), supporting the connection between the carriers +0.25 ∆α (arcsec) −80 ofALMA these observations two DIBs and of the H i. Antennae Still, the Galaxy atomic in hydrogen Carbon monox-extends wellide (CO) beyondCaltech are the available Millimeter region (Whitmore with Array detection CO map et al. of (Wilson 2014). DIB absorption, et However, al. 2000)as theyonce if λ Fig. 8. −100 present contribution. Most of the 5780 DIB absorption is found thedoresampled not characteristic cover and the binned. whole scale Unitsarea lengths are of arbitrary interest, for these and and thewere in colour particular, smaller. scale Thatis the in log- re- is: in4. a Results triangular and area discussion of 0.6 in the overlapping region of the ∼ 0 atgionarithmic a given with stretch. high location, The DIB reconstructed existenceabsorption. of white-light Therefore, H i seems image we a necessary preferred is overplotted condi- to use as −120 +0.00 system with the maximum ( 340 mÅ) at its centre. Addition- thereference Caltech with Millimeter contours in Array logarithmic CO map stretching presented of 0.5 by dex Wilson steps. et al. 4.1. Mapping the DIBs in the∼ system tion for the existence of DIB carriers, but not enough to ensure it 80 60 40 20 0 −20 −40 −60 ally, this DIB is also detected in several smaller locations in both alone.(2000). Similar The observations results have were been taken found with in our a synthesised Galaxy when beam com- of The maps showing the detections of DIBs at λ5780 and λ5797 ∆α (arcsec) galaxies. The structure of this distribution shows an uncanny re- paring300.15 4 radial00.91. A profiles resampled of these and binned species version (Puspitarini of this et CO al. map2015 is). are presented in Fig. 6. They constitute the main outcome of the × semblance to that of the obscured areas in the system, when com- Anpresentedincide exception with in the Fig. is location a8. narrow where tongue the most at [ intense10 , star2 ] formation joining the in present contribution. Most of the λ5780 DIB absorption is found 00 00 Fig. 9. Derived reddening maps for the ionised gas (top) and the stars pared with multi-band high spatial resolution with the HST (see northernthe systemThe cold nucleus occurs molecular with (Mirabel the gas overlap et distribution al. 1998; area.∼ WhileVigroux− in the absorption system,− et al. 1996). as in traced both The (bottom). The reconstructed white-light image is overplotted as refer- in a triangular area of 0.60 in the overlapping region of the warm molecular gas, as traced by the H S(2) and S(3) rotational Fig. 2 of Whitmore et∼ al. 2010), tracing not only the overlap DIBsby the have CO beenemission, detected, peaks the at VLA the northern H i 212 cm nucleus, does not is showrelatively any ence with contours in logarithmic stretching of 0.5 dex steps. area,system but with also the an maximumimportant part ( 340 of themÅ) northern at its centre. spiral arm.Addition- This lines, displays a similar spatial structure (Brandl et al. 2009). ∼ particularlystrong at the strong southern emission one and there. presents The spatial three resolution important of local the suggestsally, this DIBa relationship is also detected with the in severalextinction, smaller as has locations been found in both in maximaMolecular at roughly gas and [30 DIBs, 60 display], [30 quite, 50 different] and distributions. [35 , 45 ] VLA image, larger than00 the− width00 of00 the− tongue,00 may00 explain− 00 thegalaxies. Milky The Way structure and other of this nearby distribution galaxies shows (see e.g. an uncannyWelty et re- al. thiswhereWith eff the theect. exception two Our disks DIB of overlap detections some (Wilsonmarginal and the et DIB al. VLA 2000; detection map Zhu would close et al. still 2003;to the be 2014semblance, and referencesto that of the therein) obscured and areas will in be the explored system, in when detail com- in compatibleSmithnorthern et al.nucleus, with 2007; the we Zaragoza-Cardiel need find of no H DIBi for absorption having et al. 2014). DIB up absorption toThis our does thresh- if co- at 4.4. Attenuation and DIBs Sects.pared with4.4 and multi-band 4.5. high spatial resolution with the HST (see old in places with strong molecular gas emission. Thisi is in line this3 location there were a localised concentration of H with typ- Fig.As 2 ofreviewed Whitmore in the et introduction, al. 2010), tracing studies not mostly only in the the overlap Milky icalwithhttps://ned.ipac.caltech.edu/ size results larger for than the Milky the width Way of that the show tongue that but DIB smaller carriers than tend the It is well established that the DIB strength presents a good corre- Way, but also in nearby galaxies, show that DIBs may or may VLAto disappear resolution. in dense cores of clouds (Lan et al. 2015) - the so- lation with the interstellar extinction in the Milky Way (see e.g. notArticle be number,related withpage 6other of 12 constituents of the ISM. In that sense, called skin effect(Snow & Cohen 1974) - and that there is a lack Capitanio et al. 2017; Lan et al. 2015, and references therein) we can profit from the wealth of ancillary information about the 3ofhttps://ned.ipac.caltech.edu/ correlation between DIB strength and molecular gas (Fried- and other galaxies (e.g. Cordiner et al. 2011). Here, we obtained

A33, page 6 of 12 Article number, page 7 of 12 A. Monreal-Ibero et al.: Diffuse interstellar bands λ5780 and λ5797 in the Antennae Galaxy as seen by MUSE

VLA − HI 21 cm man et al. 2011; Welty 2014). Still, the maximum of the DIB +315 40 absorption is found in a region with mild emission in CO, and both DIBs and CO are detected in the northern spiral arm. 20 The lack of DIB detection at the location with strong CO emission coupled with the detection in the places with mild CO +235 0 + emission in Arp 244 could be explained in the context of the skin effect, as in the Milky Way. This is interpreted as being that DIB −20 formation takes place most efficiently in the outer regions of the interstellar cloud. Additionally, extinction could play a role, as +155 −40 will be discussed in Sect. 4.4. (arcsec)

∆δ −60 E(B−V) (gas) A. Monreal-Ibero et+ al.: Diffuse interstellar bands λ5780 and λ5797 in the Antennae Galaxy as seen by MUSE +1.00 −80 +75 40 VLA − HI 21 cm man et al. 2011; Welty 2014). Still, the maximum of the DIB −100 +315 20 40 absorption is found in a region with mild emission in CO, and both DIBs and CO are detected in the northern spiral arm. +0.75 −120 −5 0 + The lack of DIB detection at the location with strong CO 20 80 60 40 20 0 −20 −40 −60 emission−20 coupled with the detection in the places with mild CO ∆α +235 0 (arcsec)+ emission in Arp 244 could be explained in the context of the skin effect, as in the Milky Way. This is interpreted as being that +0.50 DIB Fig. 7. VLA H i 21 cm map (Hibbard et al. 2001) once resampled and −40 binned.−20 Units are arbitrary and the colour scale is in linear stretch. The formation (arcsec) takes place most efficiently in the outer regions of the

interstellar∆δ −60 cloud. Additionally, extinction could play a role, as reconstructed white-light image is overplotted as reference with +155 con- tours in−40 logarithmicA. stretching Monreal-Ibero of 0.5 et dex al.: steps. Diffuse interstellar bands λ5780will and λ be5797 discussed in the Antennae in Sect. Galaxy 4.4. as+ seen by MUSE +0.25

(arcsec) −80

∆δ −60 Caltech Millimeter Array − CO E(B−V) (gas) + +0.40 −100 +1.00 −8040 +75 40 −120 +0.00 −10020 20 80 60 40 20 0 −20 −40 −60 −0.25 ∆α (arcsec) +0.75 −1200 + −5 0 + E(B−V) (stars) 80 60 40 20 0 −20 −40 −60 +1.00 −20 −20 ∆α (arcsec) 40 −0.90 +0.50 Fig. 7.−40VLA H i 21 cm map (Hibbard et al. 2001) once resampled and −4020 (arcsec) binned. (arcsec) Units are arbitrary and the colour scale is in linear stretch. The +0.75 ∆δ reconstructed∆δ −60 white-light image is overplotted as reference with con- −600 + tours in logarithmic stretching of 0.5 dex+ steps. + −1.55 +0.25 −80 −20−80 Caltech Millimeter Array − CO +0.50 −100 +0.40 −100−40 40 −2.20 (arcsec) −120 +0.00 −120 ∆δ −60 20 80 60 40 20 0 −20 −40 −60 80 60 40 20 + 0 −20 −40 −60 ∆α −0.25 −80 ∆α (arcsec) +0.25 0 (arcsec)+ E(B−V) (stars) Fig.Fig. 8. CaltechCaltech Millimeter Millimeter Array Array CO CO map map (Wilson (Wilson et etal. al.2000 2000)) once once re- −100 +1.00 resampledsampled−20 and and binned. binned. Units Units are are arbitrary arbitrary and and the the colour colour scale scale is in log- log- 40 arithmicarithmic stretch. stretch. The The reconstructed reconstructed white-light white-light image image is is overplotted overplotted −0.90 as as −120 +0.00 referencereference−40 with with contours contours in in logarithmic logarithmic stretching stretching of of 0.5 0.5 dex dex steps. steps. 20 80 60 40 20 0 −20 −40 −60 (arcsec) ∆α +0.75 ∆δ −60 0 (arcsec)+ incide4.3. Molecular with the location gas and where DIBs the most+ intense star formation in −1.55 Fig. 9. Derived reddening maps for the ionised gas (top) and the stars the system−80 occurs (Mirabel et al. 1998; Vigroux et al. 1996). The (bottom−20). The reconstructed white-light image is overplotted as refer- warmALMA molecular observations gas, asof traced the Antennae by the H Galaxy2 S(2) and in Carbon S(3) rotational monox- ence with contours in logarithmic stretching of 0.5 dex steps. +0.50 lines,ide (CO)−100 displays are available a similar (Whitmore spatial structure et al. 2014 (Brandl). However, et al. 2009). they do −40 not cover the whole area of interest, and in particular, the re-

Molecular gas and DIBs display quite different distributions. (arcsec) gion with high DIB absorption. Therefore, we preferred −2.20 to use With−120 the exception of some marginal DIB detection close to the ∆δ −60 northernthe Caltech nucleus,80 Millimeter60 we find Array40 no CO20 DIB map absorption0 presented−20 up−40 by to Wilson our−60 thresh- et al. + 4.4. Attenuation and DIBs +0.25 old(2000 in). places The observations with strong∆α molecular were (arcsec) taken gas with emission. a synthesised This is beam in line of −80 3.15 4.91. A resampled and binned version of this CO map is with00 results× 00 for the Milky Way that show that DIB carriers tend It is well established that the DIB strength presents a good corre- toFig.presented disappear 8. Caltech in inFig. Millimeter dense8. cores Array of clouds CO map (Lan (Wilson et al. et 2015) al. 2000) - the once so- lation−100 with the interstellar extinction in the Milky Way (see e.g. calledresampledThe skin cold and eff molecularbinned.ect(Snow Units & gas are Cohen distribution arbitrary 1974) and - in andthe the colour that system, there scale as is is atracedin lack log- Capitanio et al. 2017; Lan et al. 2015, and references therein) ofarithmicby correlation the CO stretch. emission, between The reconstructed peaks DIB at strength the white-light northern and molecular image nucleus, is overplotted is gas relatively (Fried- as and other−120 galaxies (e.g. Cordiner et al. 2011). Here, we obtained +0.00 referencestrong at with the contours southern in onelogarithmic and presents stretching three of 0.5 important dex steps. local 80 60 40 20 0 −20 −40 −60 , , , maxima at roughly [3000 6000], [3000 5000] and [3500 4500] ∆α (arcsec) Article number, page 7 of 12 where the two disks overlap− (Wilson et al.− 2000; Zhu et al.− 2003; incide with the location where the most intense star formation in Smith et al. 2007; Zaragoza-Cardiel et al. 2014). This does coin- Derived reddening maps for the ionised gas (top) and the stars the system occurs (Mirabel et al. 1998; Vigroux et al. 1996). The Fig. 9. Derived reddening maps for the ionised gas (top) and the stars cide with the location where the most intense star formation in (bottom). The reconstructed white-light image is overplotted as refer- warm molecular gas, as traced by the H S(2) and S(3) rotational (bottom). The reconstructed white-light image is overplotted as refer- the system occurs (Mirabel et al. 1998; 2Vigroux et al. 1996). The ence with contours in logarithmic stretching of 0.5 dex steps. lines, displays a similar spatial structure (Brandl et al. 2009). warm molecular gas, as traced by the H S(2) and S(3) rotational Molecular gas and DIBs display quite2 different distributions. lines, displays a similar spatial structure (Brandl et al. 2009). With the exception of some marginal DIB detection close to the Molecular gas and DIBs display quite different distributions. northern nucleus, we find no DIB absorption up to our thresh- 4.4. Attenuation and DIBs With the exception of some marginal DIB detection close to the 4.4. Attenuation and DIBs old in places with strong molecular gas emission. This is in line northern nucleus, we find no DIB absorption up to our thresh- It is well established that the DIB strength presents a good corre- with results for the Milky Way that show that DIB carriers tend It is well established that the DIB strength presents a good corre- old in places with strong molecular gas emission. This is in line lation with the interstellar extinction in the Milky Way (see e.g. to disappear in dense cores of clouds (Lan et al. 2015) - the so- lation with the interstellar extinction in the Milky Way (see e.g. with results for the Milky Way that show that DIB carriers tend Capitanio et al. 2017; Lan et al. 2015, and references therein) called skin effect(Snow & Cohen 1974) - and that there is a lack Capitanio et al. 2017; Lan et al. 2015, and references therein) to disappear in dense cores of clouds (Lan et al. 2015) – the and other galaxies (e.g. Cordiner et al. 2011). Here, we obtained of correlation between DIB strength and molecular gas (Fried- and other galaxies (e.g. Cordiner et al. 2011). Here, we obtained so-called skin effect (Snow & Cohen 1974) – and that there is a total number of independent measurements of DIB strength a lack of correlation between DIB strength and molecular gas within the system of >200 for the λ5780Article DIB number, and > page100 for7 of the 12 (Friedman et al. 2011; Welty 2014). Still, the maximum of the λ5797 DIB. This is comparable to the number of lines of sight DIB absorption is found in a region with mild emission in CO, typically used in many Galactic studies. Thus, we are in an opti- and both DIBs and CO are detected in the northern spiral arm. mal position to explore whether a similar relationship exists also The lack of DIB detection at the location with strong CO in an environment experiencing a star formation as extreme as emission coupled with the detection in the places with mild CO that found in starburst galaxies. emission in Arp 244 could be explained in the context of the skin We present two reddening maps for the Antennae Galaxy effect, as in the Milky Way. This is interpreted as being that DIB in Fig.9. They are not corrected for Galactic extinction. formation takes place most efficiently in the outer regions of the In the direction of the Antennae, this is AV = 0.127 interstellar cloud. Additionally, extinction could play a role, as (Schlafly & Finkbeiner 2011). Assuming a selective redden- will be discussed in Sect. 4.4. ing of RV = 3.1 (Rieke & Lebofsky 1985), this implies an

A33, page 7 of 12 A&A proofs: manuscript no. aanda a total number of independent measurements of DIB strength stellar population as seen in the optical is subject of relatively within the system of > 200 for the λ5780 DIB and > 100 for the low attenuation while that derived for the ionised gas is rela- λ5797 DIB. This is comparable to the number of lines of sight tively high. This can be understood in a context where the optical typically used in many Galactic studies. Thus, we are in an opti- stellar continuum is blind to the very heart of the overlap region mal position to explore whether a similar relationship exists also where most of the massive star formation takes place. E(B V) in an environment experiencing a star formation as extreme as values provided by STARLIGHT and based on the stellar contin-− that found in starburst galaxies. A&A 615, A33uum (2018) trace only the most superficial layers in this area. Then, in these layers the concentration of DIBs carriers should be low, 500 500 N points = 227 N points = 230 too.We Togetherwill come with back the to previously this point mentioned when discussing skin effect, the this relation may Pearson's r : 0.689 Pearson's r : 0.626 ) ) explain the lack of DIB absorption in this area. We will come Å 400 Spearman's r : 0.632 Å 400 Spearman's r : 0.663 between DIB absorption and emission in the mid-infrared band m m ( (

) 300 ) 300 back(see Sect. to this 4.6 point). when discussing the relation between DIB ab- 0 0 8 8

7 7 sorption and emission in the mid-infrared band (see Sect. 4.6).

5 5 The relationship between DIBs and attenuation is further ex- 200 200 ( ( λ

W W plored in Fig. 10. The upper row shows the results for the 5780 E 100 E 100 DIB.The In relationship the locations between where theDIBs DIB and was attenuation detected, is it further correlates ex- plored in Fig. 10. The upper row shows the results for the λ5780 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 equally well with both the stellar and ionised gas attenuation, in E(B-V) E(B-V) DIB.the sense In the that locations both the where Pearson the and DIB the was Spearman detected, correlation it correlates co- 200 200 equally well with both the stellar and ionised gas attenuation, N points = 119 N points = 120 efficients are comparable. The correlation is clear and significant 175 Pearson's r : 0.423 175 Pearson's r : 0.388

) ) in the sense that both the Pearson and the Spearman correlation

Å Spearman's r : 0.383 Å Spearman's r : 0.436 according to a Student’s t-test, although the coefficients are not 150 150 m m ffi ( ( coe cients are comparable. The correlation is clear and signifi-

) 125 ) 125 as high as those determined for our Galaxy (typically using much 7 7

9 9 cant according to a Student’s t-test, although the coefficients are 7 100 7 100 higher spatial and spectral resolution, e.g. Friedman et al. 2011; 5 5

( 75 ( 75 notPuspitarini as high etas al. those 2013 determined). Actually, for the our difference Galaxy in (typically spectral resolu- using W W E 50 E 50 muchtion can higher at least spatial partially and explainspectral the resolution, lower strength e.g. Friedman of the correla- et al. 25 25 2011; Puspitarini et al. 2013). Actually, the difference in spec- 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 tion for the Antennae Galaxy. The λ5780 is blended with another E(B-V) E(B-V) tralbroader resolution DIB at canλ5779 at least (Jenniskens partially explain & Desert the 1994 lower). Here, strength since of λ thethe correlation spectral resolution for the Antennaeis relatively Galaxy. low, we The did not5780 make is blended any at- Fig.Fig. 10. 10. ComparisonComparison between between the the strength strength of of the the DIBs DIBs at at λλ57805780 ( (toptop)) λ withtempt another to separate broader both DIB features, at 5779 and thus(Jenniskens variation & in Desert their relative 1994). andand at at λλ57975797 ( (bottombottom)) and and the the reddening reddening of of the the ionised ionised gas gas ( (leftleft)) and and Here, since the spectral resolution is relatively low, we did not the stars (right). Uncertainties for the gas reddening range typically be- contributions to the measured equivalent width can have an im- the stars (right). Uncertainties for the gas reddening range typically be- make any attempt to separate both features, and thus variation in tweentween 0.06 0.06 and and 0.14 0.14 dex. dex. Uncertainties Uncertainties for for the theλλ57805780 and andλλ57975797 equiv- equiv- pact on the robustness of the E(B V)–EW(λ5780) correlation. their relative contributions to the measured− equivalent width can alentalent widths widths were were 5050 and and 4040 mÅ, mÅ, respectively respectively (see (see Sec. Sect. 3.3). 3.3). A relationship between the λ5797 DIB and attenuation is not ∼∼ ∼∼ have an impact on the robustness of the E(B V) - EW(λ5780) that clear from the present data. Correlation coe− fficients are low, We present two reddening maps for the Antennae Galaxy correlation.and according to them only about 20–25% of the variation in the inE( FigureB V) 9.= They0.041. are In not galaxies corrected where for one Galactic cannot extinction. resolve in- In EW(λ5797) is somehow relatedλ to the amount of reddening. This − A relationship between the 5797 DIB and attenuation is not thedividual direction stars, of the the mostAntennae, similar this accessible is AV = 0 magnitude.127 (Schlafly to the & thatis in clear clear from contrast the presentwith results data. for Correlation the Milky coe Way,fficients for which are low, this Finkbeinerextinction is 2011). luminosity-weighted Assuming a selective attenuation, reddening that of canRV present= 3.1 andDIB according is an even to better them proxyonly about for the 20 extinction25% of thethan variation the one in at (Rieke &ff Lebofsky 1985), this impliesff an E(B V)=0.041. In λ5780 (Ensor et al. 2017). However, these− low correlation coef- some di erences due to transfer e ects and− scattering into the EW(λ5797) is somehow related to the amount of reddening. galaxiesthe line where of sight. one Still, cannot these resolve differences individual in stars,the attenuation the most Thisficients is in are clear largely contrast weighted with results by a series for the of Milky measurements Way, for which along similarand extinction accessible curves magnitude are mostly to the found extinction in the isultraviolet luminosity- be- thisseveral DIB lines is an of even sight better with proxy EW( forλ5797) the extinction of only about than twothe one times at weighteding essentially attenuation, the same that in can the present optical some (Gordon differences et al. due2003 to), λthe5780 estimated (Ensor etuncertainty. al. 2017). Most However, of the these data low points correlation in fact lie coef- be- transferand more effects specifically and scattering in the into wavelength the line of range sight. used Still, in these this ficientstween 40 are and largely 100 weighted mÅ. Thus by we a series are mostly of measurements sampling a along range diwork.fferences Thus in we the will attenuation assume here and extinction that extinction curves laws are derived mostly severalin equivalent lines of widths sight comparable with EW(λ5797) to the of estimated only about uncertainties, two times foundfor our in Galaxy the ultraviolet can be being used essentially to discuss the attenuation same in the in the optical op- theimplying estimated a meagre uncertainty. contrast Most to detect of the a data solid points correlation. in fact lie More- be- (Gordon et al. 2003), and more specifically in the wavelength tical within the system. The map in the upper panel was tweenover, this 40 and DIB 100 is much mÅ. Thus narrower we are than mostly that at samplingλ5780, and a range is also in range used in this work. Thus we will assume here that extinc- derived using the extinction curve presented by Fluks et al. equivalentblended with widths a broader comparable DIB at toλ5795. the estimated In that sense uncertainties, it is more sen- im- tion laws derived for our Galaxy can be used to discuss atten- (1994) and assuming an intrinsic Balmer emission-line ratio plyingsitive to a meagrethe spectral contrast resolution to detect of thea solid instrument correlation. than Moreover, the λ5780 uation in the optical within the system. The map in the upper Hα/Hβ = 2.86 for a case B approximation and Te = 10 000 K thisDIB. DIB is much narrower than that at λ5780, and is also blended panel was derived using the extinction curve presented by Fluks (Osterbrock & Ferland 2006). As such, it represents the atten- withOn a broader the positive DIB at side,λ5795. when In that looking sense at it is high more EW( sensitiveλ5797)s, to et al. (1994) and assuming an intrinsic Balmer emission-line ra- uation towards the ionised gas (and young stellar population). thethere spectral is a clear resolution tendency of the in instrument the sense thanthat lines the λ5780 of sight DIB. with tioThe Hα one/Hβ in= 2.86 the forlower a case panel B approximation was provided and by T STARLIGHT.e = 10 000 K larger equivalent width also have higher gas extinction. This is (OsterbrockIt represents & theFerland attenuation 2006). As of such, the it overall represents stellar the attenua- popula- encouragingOn the positive and calls side, for a when similar looking experiment at high (in this EW( orλ5797)s, another tiontion towards at optical the wavelengths. ionised gas (and In both young cases, stellar attenuation population). is high The theregalaxy) is witha clear an tendency IFS similar into the MUSE sense butthat with lines higher of sight spectral with onein the in the overlap lower panel region, was being provided higher by STARLIGHT when using. It the represents Hα to largerresolution. equivalent This is width exactly also what have the higher spectrographs gas extinction. foreseen This for is theHβ attenuationratio. This of has the already overall been stellar seen population in other at starbursts optical wave- and encouragingthe future ESO and ELT,calls for HARMONI, a similar experiment and MOSAIC, (in this will or another be able lengths.can be In explained both cases, if attenuation the dust has is high a larger in the covering overlap region, factor galaxy)to provide. with In an the IFS meantime, similar to instruments MUSE but such with as higher MEGARA spectral on beingfor ionised higher gas when than using for the the Hα starsto H (e.g.β ratio. Calzetti This et has al. already 1997; resolution.GTC could This also be is usedexactly to obtain what the data spectrographs with reduceduncertainties foreseen for beenMonreal-Ibero seen in other et al.starbursts 2010). and can be explained if the dust has theand future covering ESO a larger ELT, range HARMONI, of equivalent and MOSAIC, widths. will be able a largerWhen covering comparing factor both for maps ionised with gas those than in for Fig. the6 a stars clear (e.g. re- to provide.To find In out the whether meantime, these instruments measurements such support as MEGARA a putative on Calzettisult arises: et al. the 1997; overall Monreal-Ibero structure is comparable et al. 2010). and DIBs are de- GTCcorrelation could alsobetween be used this to DIB obtain and dataextinction, with reduced we can uncertainties compare our tectedWhen there comparing where attenuation both maps iswith high. those The in highestFig. 6 a clearattenuation result andresults covering with those a larger found range for of other equivalent galaxies, widths. that sample a larger arises:for both, the ionised overall gas structure and overall is comparable stellar population, and DIBs does are detectedcoincide range of equivalent widths. This is the subject of the discussion therewith thewhere location attenuation with largest is high. DIB The absorptionhighest attenuation (at [25 for00, both,2500] in Sect.To find 4.5. out whether these measurements support a putative ionisedin our relative gas and coordinates overall stellar system). population, Still, does in the coincide area∼ with with− high- the correlation between this DIB and extinction, we can compare our location with largest DIB absorption (at [2500, 2500] in our results with those found for other galaxies, that sample a larger est concentration of molecular gas (and no∼ DIB detection),− the 4.5. Attenuation and DIBs here and elsewhere relativeoverall stellar coordinates population system). as seen Still, in in the the optical area with is subject highest of con- rel- range of equivalent widths. This is the subject of the discussion centrationatively low of attenuation molecular while gas (and that no derived DIB detection), for the ionised the overall gas is inIn Sect. this section, 4.5. we discuss how the relationships between red- relatively high. This can be understood in a context where the dening and strength of the DIBs found in the previous section Articleoptical number, stellar pagecontinuum 8 of 12 is blind to the very heart of the over- compares with those found in other galaxies. lap region where most of the massive star formation takes place. This is summarised in Fig. 11 for the λ5780 DIB. The fig- E(B V) values provided by STARLIGHT and based on the ure is similar to Fig. 4 by Monreal-Ibero et al.(2015b), but with stellar− continuum trace only the most superficial layers in this the new data for the Antennae Galaxy overplotted. It illustrates area. Then, in these layers the concentration of DIBs carriers beautifully the enormous potential of IFSs for extragalactic DIB should be low, too. Together with the previously mentioned skin research: passing from a few tens to a few hundreds of extra- effect, this may explain the lack of DIB absorption in this area. galactic measurements of the λ5780 DIB was possible by adding

A33, page 8 of 12 A. Monreal-Ibero et al.: Diffuse interstellar bands λ5780 and λ5797 in the Antennae Galaxy as seen by MUSE

Fig. 11. Relation between the strength of the λ5780 DIB and the red- dening for the Antennae Galaxy (green circles for the gas extinction), and other galaxies. For the Antennae Galaxy data, a Galactic reddening E(B V)gal = 0.041 has been subtracted. Regarding the data for other galaxies,− we included two sets of observations for Galactic targets (blue triangles; filled: Puspitarini et al. 2013, open: Friedman et al. 2011), several sets of data for galaxies in the Local Group (inverted salmon triangles: Magellanic Clouds, Welty et al. 2006; filled burgundy “x”: M 31, Cordiner et al. 2011, open burgundy “x”: M 33, Cordiner et al. 2008), a set of measurements through dusty starbursts (black diamonds, Heckman & Lehnert 2000), some measurements in the lines of sight of Fig. 13. Spitzer/IRAC image in the 8 µm band (Smith et al. 2007) once supernovae (filled blue crosses, Phillips et al. 2013), including one in resampled and binned. Units are arbitrary and the colour scale is in log- M 82 (coral pink pentagon, Welty et al. 2014) and our previous results arithmic stretch. The reconstructed white-light image is overplotted as for AM 1353-272 B (red circles and blue stars, Monreal-Ibero et al. reference with contours in logarithmic stretching of 0.5 dex steps. 2015b).

when a larger range in equivalent widths is considered, the corre- lation between this DIB and reddening is unveiled. Interestingly enough, the separation between data for the Magellanic Clouds and for galaxies with solar or solar-like metallicities is not as neat for this DIB as for the λ5780 DIB. This difference in be- haviour is in line with the idea of different carriers for these two absorption bands and in harmony with results that show, for ex- ample, the different reactions of these two DIBs to the radiation field (e.g. Vos et al. 2011; Cami et al. 1997).

4.6. Mid-IR emission and DIBs Neutral and/or ionised PAHs are among the favourite candidates Fig. 12. Same as Fig. 11 but for the λ5797 DIB. We note that results of to be (some of) the DIB carriers (Salama et al. 1996). Several Monreal-Ibero et al.(2015b) and Phillips et al.(2013) are not included, strong emission bands in the mid-IR at, for example, 3.3, 6.2, since these authors did not measure the λ5797 DIB. 7.7, and 8.6 µm, the so-called unidentified infrared emission bands (UIBs), are also believed to be caused by PAHs and re- lated species (e.g. Rigopoulou et al. 1999; Draine & Li 2007)6. only one set of observations. As with AM 1353-272 B, most of our Thus, in the context of gaining insight into the nature of the car- measurements for Arp 244 fall very nicely in the relation defined riers, a comparison between our DIB maps and the emission at by our Galaxy, M 31, and M 33 all of them spiral galaxies with these wavelengths is an important test. solar or near-solar metallicities (e.g. Toribio San Cipriano et al. Figure 13 shows the Spitzer/IRAC image in the 8.0 µm 4 2016; Lardo et al. 2015; Sanders et al. 2012) . and separated from band downloaded from the NED (Smith et al. 2007, see also the relation defined by the Magellanic Clouds5. Thus, our results Wang et al. 2004). The original spatial resolution was 100.6. As support the existence of a second parameter (e.g. metallicity, the for the H i 21 cm line and CO maps, it was re-sampled∼ and characteristics of the radiation field) modulating the E(B V)– − binned to match our tessellation and ease the comparison with EW(λ5780) relation. Similar observations to those presented here the DIB maps. This image basically reflects emission from the for additional dwarf(-ish) galaxies at sub-solar metallicities (for strong 7.7 and 8.6 µm bands, with minor (i.e. <5%) stellar contri- example) would be instrumental to further discuss the relative role bution (Wang et al. 2004). A similar image for the 5.8 µm band of the possible second parameter. – that contains the strong 6.2 µm UIB – is also available. We did Figure 12 contains a similar relation to that in Fig. 11 but not include it here since it displays a similar morphology to the for the λ5797 DIB. Even if the data uncertainty is large, here one at 8.0 µm. Additionally, Brandl et al.(2009), by scanning we see that this DIB actually falls in the locus defined by mea- with the Spitzer/IRS, obtained spectral maps for the 8.6 µm and surements in other galaxies (including the Milky Way) and that 6 UIBs are often directly referred in the literature as the PAH bands 4 Here we assume 12 + log(O/H) = 8.69 (Asplund et al. 2009). or PAH emission. Here, we reserve the acronym UIBs to refer to the 5 Metallicities for the LMC and SMC are 12 + log(O/H) = 8.35 and spectroscopic features and the acronym PAHs to refer to the molecules 8.03 (Russell & Dopita 1992). themselves, to avoid ambiguities.

A33, page 9 of 12 A&A 615, A33 (2018)

11.3 µm UIBs (see their Fig. 4). All these maps display roughly 6. The λ5797 DIB does not strongly correlate with the ex- the same structure: strong emission is seen in the two nuclei of tinction. We attribute this mainly to a lack of contrast for the galaxies, the outer spiral arm of NGC 4038, and the overlap the covered equivalent widths and estimated uncertainties. region, with an irregular patchy structure. Still, data points with high EW(λ5797) indicate that this We note that there is a marked difference between the struc- connection may actually exist. ture of the UIBs (presumably caused by PAHs) and our DIB 7. When comparing our results with those for the EW(λ5780)– maps in Fig.6. While UIBs are strong over the whole overlap E(B V) relation for other galaxies, the λ5780 DIB data falls region, where atomic and especially molecular gas are strong, in the− sequence defined by other solar or solar-like spiral DIBs are detected only there where the H i concentration is the galaxies and far from the locus defined by the Magellanic highest and CO emission is moderate. This result does not nec- Clouds. The compilation of sets of data similar to these for essarily imply that PAHs or (PAH-related molecules) can be re- an ensemble of galaxies with different metallicities would jected as possible carriers of the λ5780 and λ5797 DIBs. As we open the possibility of establishing relations of the kind of, discussed in Sect. 4.4, in the area of high CO emission, optical for example, E(B V) = f (EW(DIB), Z). This would allow and mid-infrared are actually sampling different depths along the the use of DIBs as− a proxy for the extinction in the absence line of sight. Should these two DIBs and the UIBs share related of any other available information. carriers, it is therefore reasonable to expect that their spectral sig- 8. When comparing our results with the EW(λ5797)–E(B V) natures are inaccessible in the optical and can only be detected in relation for other galaxies, the correlation between these− two the mid-infrared. This would be a situation similar to that for the quantities is revealed. The influence of a second parameter stars: most of the recent star formation is hidden in the optical in this relation seems less marked here as for the case of the (Whitmore et al. 2010) but revealed in the near and mid-infrared EW(λ5780)–E(B V) relation. (Brandl et al. 2005; Wang et al. 2004). In that context, it would 9. The distributions of− UIBs in the mid-IR and DIBs are similar be interesting to see whether infrared DIBs, less sensible to ex- but with some differences. The largest ones are in the area tinction effects are found in this area. of the highest CO emission. We preferentially attribute these differences to extinction effects rather than necessarily implying a very different nature of the UIB and DIB 5. Conclusions carriers. This is the second of a series of papers aiming at a detailed study The present work demonstrates how IFSs are an efficient tool to of the Antennae Galaxy with MUSE. At the same time, it is the collect in one shot and in a relatively short time plenty of infor- continuation of an experiment devoted to explore the potential mation about DIBs in galaxies other than our own. With enough of using IFSs in extragalactic DIB research. In this contribution, spectral resolution, the working philosophy could also be applied we used high S/N spectra of 1200 lines of sight to measure the to other ISM absorption features. The itemised conclusions of absorption strength of the well-known∼ λ5780 and λ5797 DIBs. this work illustrate how the construction of 2D maps for DIBs We examined the correlations between them and the extinction can be used to better understand the relationship between these in the system, and how their spatial distributions compare with features and other components of the ISM in galaxies. In Sect.4, that of other components of the ISM. Our main findings are: we showed that DIB distributions present good – but not perfect – 1. To our knowledge, we have derived the first spatially re- correlations with other constituents of the ISM and some possible solved maps for the DIBs at λ5780 and λ5797 in galaxies explanations for these differences were discussed. An open issue outside the Local Group. The detection area was largest for in the comparisons presented in this work is the consequence of the strongest of the two DIBs (at λ5780) and covered a region the – surely existing – projection effects. Given the violent nature 2 of 0.60, corresponding to a linear area of 25 kpc . of the interaction in the system, there is a strong possibility that 2. Both∼ DIBs correlate with atomic hydrogen in∼ the sense that the stars and gas are not well mixed, which in turns could also they were detected in the area with largest amount of this partially explain the different distributions mapped for the DIBs, gas phase. However, typical length scales for these species E(B V), H i, and UIBs. From the kinematic point of view, the are different, with atomic hydrogen extending well beyond individual− galaxies still display relatively regular velocity fields the area presenting DIB absorption. Thus, the existence of (see Paper I, for the ionised gas, Weilbacher et al., in prep. for the atomic hydrogen seems a necessary condition but not suffi- stars), with the exception precisely in the overlap region where cient for the existence of DIBs. attenuation high. Since i) this is the first paper that uses IFS-based 3. DIBs and molecular gas present only mildly similar distribu- data to map DIBs in a situation of non-resolved stellar population, tions. This is in accord with results for the Milky Way that ii) a merging system is not the easiest case to start with a thor- indicates a lack of correlation between molecular gas and the ough discussion about projection effects, we have focused here λ5780 and λ5797 DIBs. DIBs are found in areas with mod- on the methodology to extract the information and on the possi- erately CO emission but they are absent in those with the bilities that it can bring, in a more generic manner, to emphasise highest CO emission. Both, the skin effect and the high ex- the potential of the technique. Specifically, we focused on two of tinction towards the areas with the highest CO emission can the strongest DIBs and their relation with extinction, atomic and partially explain this. molecular gas, and the PAHs as traced by the UIBs in the mid- 4. The overall DIB and extinction distributions are comparable: infrared. Similarly, one could explore the association (or lack of) the locations for the highest DIB absorption and highest with other components and characteristics of the ISM. For ex- extinction do coincide. However, large extinction for the ample, one may wonder how these (and other) DIBs depend on ionised gas is seen also there where the concentration of the dust-to-gas ratio, the radiation field, or the gas metallicity molecular gas is maximal and no DIB absorption has been and relative chemical abundances. The uncertainty of the current detected. data prevent us from exploring these relations. However higher 5. Within the system, the DIB λ5780 correlates well with both quality data (i.e. deeper and at higher spatial resolution as those the gas and stellar extinction, similar to what happens with that will be provided by, for example, HARMONI at the E-ELT) this DIB in the Milky Way. will be instrumental for these purposes. Learning about these

A33, page 10 of 12 A. Monreal-Ibero et al.: Diffuse interstellar bands λ5780 and λ5797 in the Antennae Galaxy as seen by MUSE relationships will certainly be key to evaluate the diagnostic po- Gordon, K. D., Clayton, G. C., Misselt, K. A., Landolt, A. U., & Wolff, M. J. tential of these features and will help to put constrains on the na- 2003, ApJ, 594, 279 ture of the carriers. Hamano, S., Kobayashi, N., Kondo, S., et al. 2016, ApJ, 821, 42 Heckman, T. M., & Lehnert, M. D. 2000, ApJ, 537, 690 Heger, M. L. 1922, Lick Obs. Bull., 10, 141 Acknowledgements. We thank Christine Wilson for making the Caltech Mil- Herbig, G. H. 1993, ApJ, 407, 142 limeter Array CO map available to us. We also thank the referee for valu- Herbig, G. H. 1995, ARA&A, 33, 19 able and supportive comments that have significantly helped us to improve Hibbard, J. E., van der Hulst, J. M., Barnes, J. E., & Rich, R. M. 2001, AJ, 122, the first submitted version of this paper. AMI acknowledges support from the 2969 Spanish MINECO through project AYA2015-68217-P. PMW received support Hobbs, L. M., York, D. G., Thorburn, J. A., et al. 2009, ApJ, 705, 32 through BMBF Verbundforschung (projects MUSE-AO, grant 05A14BAC, and Hunter, J. D. 2007, Comput. Sci. Eng., 9, 90 MUSE-NFM, grant 05A17BAA). Based on observations collected at the Eu- Hutchings, J. B. 1966, MNRAS, 131, 299 ropean Organisation for Astronomical Research in the Southern Hemisphere Iglesias-Groth, S. 2007, ApJ, 661, L167 under ESO programmes 095.B-0042, 096.B-0017, and 097.B-0346. This re- Jenniskens, P., & Desert, F.-X. 1994, A&AS, 106, 39 search has made use of the NASA/IPAC Extragalactic Database (NED) which Joblin, C., D’Hendecourt, L., Leger, A., & Maillard, J. P. 1990, Nature, 346, 729 is operated by the Jet Propulsion Laboratory, California Institute of Technology, Jones, E., Oliphant, T., Peterson, P., et al. 2001, SciPy: Open Source Scientific under contract with the National Aeronautics and Space Administration. This Tools for Python paper uses the plotting package jmaplot developed by Jesús Maíz-Apellániz7. Junkkarinen, V. T., Cohen, R. D., Beaver, E. A., et al. 2004, ApJ, 614, 658 We are also grateful to the communities who developed the many python Kos, J. 2017, MNRAS, 468, 4255 packages used in this research, such MPDAF (Piqueras et al. 2017), Astropy Kos, J., Zwitter, T., Grebel, E. K., et al. 2013, ApJ, 778, 86 (Astropy Collaboration et al. 2013), numpy (Walt et al. 2011), scipy (Jones et al. Krełowski, J., Snow, T. P., Seab, C. G., & Papaj, J. 1992, MNRAS, 258, 693 2001) and matplotlib (Hunter 2007). Krełowski, J., Galazutdinov, G., & Kołos, R. 2011, ApJ, 735, 124 Kroto, H. 1988, Science, 242, 1139 Lan, T.-W., Ménard, B., & Zhu, G. 2015, MNRAS, 452, 3629 Lardo, C., Davies, B., Kudritzki, R.-P., et al. 2015, ApJ, 812, 160 References Leger, A., & D’Hendecourt, L. 1985, A&A, 146, 81 Asplund, M., Grevesse, N., Sauval, A. J., & Scott, P. 2009, ARA&A, 47, 481 Maier, J. P., Walker, G. A. H., & Bohlender, D. A. 2004, ApJ, 602, 286 Astropy Collaboration, Robitaille, T. P., Tollerud, E. J., et al. 2013, A&A, 558, Markwardt, C. B. 2009, in Astronomical Data Analysis Software and Systems A33 XVIII, eds. D. A. Bohlender, D. Durand, & P. Dowler, ASP Conf. Ser., 411, Bacon, R., Accardo, M., Adjali, L., et al. 2012, The Messenger, 147, 4 251 Bailey, M., van Loon, J. T., Sarre, P. J., & Beckman, J. E. 2015, MNRAS, 454, Merrill, P. W. 1936, ApJ, 83, 126 4013 Mirabel, I. F., Vigroux, L., Charmandaris, V., et al. 1998, A&A, 333, L1 Baron, D., Poznanski, D., Watson, D., et al. 2015, MNRAS, 451, 332 Monreal-Ibero, A., Vílchez, J. M., Walsh, J. R., & Muñoz-Tuñón C. 2010, A&A, Bhatt, N. H., & Cami, J. 2015, ApJS, 216, 22 517, A27 Blades, J. C., & Madore, B. F. 1979, A&A, 71, 359 Monreal-Ibero, A., Lallement, R., Puspitarini, L., Bonifacio, P., & Monaco, L. Brandl, B. R., Clark, D. M., Eikenberry, S. S., et al. 2005, ApJ, 635, 280 2015a, Mem. Soc. Astron. It., 86, 527 Brandl, B. R., Snijders, L., den Brok, M., et al. 2009, ApJ, 699, 1982 Monreal-Ibero, A., Weilbacher, P. M., Wendt, M., et al. 2015b, A&A, 576, L3 Bruzual, G., & Charlot, S. 2003, MNRAS, 344, 1000 Motylewski, T., Linnartz, H., Vaizert, O., et al. 2000, ApJ, 531, 312 Calzetti, D., Meurer, G. R., Bohlin, R. C., et al. 1997, AJ, 114, 1834 Osterbrock, D. E., & Ferland, G. J. 2006, Astrophysics of Gaseous Nebulae and Cami, J., Sonnentrucker, P., Ehrenfreund, P., & Foing, B. H. 1997, A&A, 326, Active Galactic Nuclei, (Sausalito, CA: University Science Books) 822 Phillips, M. M., Simon, J. D., Morrell, N., et al. 2013, ApJ, 779, 38 Campbell, E. K., Holz, M., Gerlich, D., & Maier, J. P. 2015, Nature, 523, 322 Piqueras, L., Conseil, S., Shepherd, M., et al. 2017, Astronomical Data Analysis Capitanio, L., Lallement, R., Vergely, J. L., Elyajouri, M., & Monreal-Ibero, A. Software and Systems XXVI, submitted, [arXiv:1710.03554] 2017, A&A, 606, A65 Puspitarini, L., Lallement, R., & Chen, H.-C. 2013, A&A, 555, A25 Cappellari, M., & Copin, Y. 2003, MNRAS, 342, 345 Puspitarini, L., Lallement, R., Babusiaux, C., et al. 2015, A&A, 573, A35 Cardelli, J. A., Clayton, G. C., & Mathis, J. S. 1989, ApJ, 345, 245 Rieke, G. H., & Lebofsky, M. J. 1985, ApJ, 288, 618 Cid Fernandes, R., Mateus, A., Sodré, L., Stasinska,´ G., & Gomes, J. M. 2005, Rigopoulou, D., Spoon, H. W. W., Genzel, R., et al. 1999, AJ, 118, 2625 MNRAS, 358, 363 Ritchey, A. M., & Wallerstein, G. 2015, PASP, 127, 223 Cid Fernandes, R., Schoenell, W., Gomes, J. M., et al. 2009, Rev. Mex. Astron. Russell, S. C., & Dopita, M. A. 1992, ApJ, 384, 508 Astrofis. Conf. Ser., 35, 127 Salama, F., Bakes, E. L. O., Allamandola, L. J., & Tielens, A. G. G. M. 1996, Cordiner, M. A. 2014, in The Diffuse Interstellar Bands, eds. J. Cami, & N. L. J. ApJ, 458, 621 Cox, IAU Symp., 297, 41 Sanders, D. B., Mazzarella, J. M., Kim, D.-C., Surace, J. A., & Soifer, B. T. Cordiner, M. A., Cox, N. L. J., Trundle, C., et al. 2008, A&A, 480, L13 2003, AJ, 126, 1607 Cordiner, M. A., Cox, N. L. J., Evans, C. J., et al. 2011, ApJ, 726, 39 Sanders, N. E., Caldwell, N., McDowell, J., & Harding, P. 2012, ApJ, 758, 133 Cordiner, M. A., Fossey, S. J., Smith, A. M., & Sarre, P. J. 2013, ApJ, 764, L10 Sarre, P. J. 2006, J. Mol. Spectr., 238, 1 Cox, N. L. J., & Patat, F. 2008, A&A, 485, L9 Sassara, A., Zerza, G., Chergui, M., & Leach, S. 2001, ApJS, 135, 263 Cox, N. L. J., & Spaans, M. 2006, A&A, 451, 973 Schlafly, E. F., & Finkbeiner, D. P. 2011, ApJ, 737, 103 Cox, N. L. J., Cordiner, M. A., Cami, J., et al. 2006, A&A, 447, 991 Schweizer, F., Burns, C. R., Madore, B. F., et al. 2008, AJ, 136, 1482 Cox, N. L. J., Cordiner, M. A., Ehrenfreund, P., et al. 2007, A&A, 470, 941 Smith, B. J., Struck, C., Hancock, M., et al. 2007, AJ, 133, 791 Cox, N. L. J., Cami, J., Kaper, L., et al. 2014, A&A, 569, A117 Snow, T. P. 2014, in The Diffuse Interstellar Bands, eds. J. Cami, & N. L. J. Cox, Cox, N., Cami, J., Farhang, A., et al. 2017, A&A, 606, A76 IAU Symp., 297, 3 D’Odorico, S., di Serego Alighieri, S., Pettini, M., et al. 1989, A&A, 215, 21 Snow, T. P. Jr., & Cohen, J. G. 1974, ApJ, 194, 313 Draine, B. T., & Li, A. 2007, ApJ, 657, 810 Sollerman, J., Cox, N., Mattila, S., et al. 2005, A&A, 429, 559 Ehrenfreund, P., Cami, J., Jiménez-Vicente, J., et al. 2002, ApJ, 576, L117 Srianand, R., Gupta, N., Rahmani, H., et al. 2013, MNRAS, 428, 2198 Elyajouri, M., Lallement, R., Monreal-Ibero, A., Capitanio, L., & Cox, N. L. J. Thöne, C. C., Michałowski, M. J., Leloudas, G., et al. 2009, ApJ, 698, 1307 2017, A&A, 600, A129 Toribio San Cipriano, L., García-Rojas, J., Esteban, C., Bresolin, F., & Peimbert, Ensor, T., Cami, J., Bhatt, N. H., & Soddu, A. 2017, ApJ, 836, 162 M. 2016, MNRAS, 458, 1866 Fluks, M. A., Plez, B., The, P. S., et al. 1994, A&AS, 105, 311 van der Zwet, G. P., & Allamandola, L. J. 1985, A&A, 146, 76 Friedman, S. D., York, D. G., McCall, B. J., et al. 2011, ApJ, 727, 33 van Loon, J. T., Smith, K. T., McDonald, I., et al. 2009, MNRAS, 399, 195 Galazutdinov, G. A., & Krełowski, J. 2017, Acta Astron., 67, 159 van Loon, J. T., Bailey, M., Tatton, B. L., et al. 2013, A&A, 550, A108 Geballe, T. R., Najarro, F., Figer, D. F., Schlegelmilch, B. W., & de La Fuente, Vigroux, L., Mirabel, F., Altieri, B., et al. 1996, A&A, 315, L93 D. 2011, Nature, 479, 200 Vos, D. A. I., Cox, N. L. J., Kaper, L., Spaans, M., & Ehrenfreund, P. 2011, Girardi, L., Bressan, A., Bertelli, G., & Chiosi, C. 2000, A&AS, 141, 371 A&A, 533, A129 Walker, G. A. H., Bohlender, D. A., Maier, J. P., & Campbell, E. K. 2015, ApJ, 812, L8 Walt, S. v. d., Colbert, S. C., & Varoquaux, G. 2011, Comput. Sci. Eng., 13, 22 7 http://jmaiz.iaa.es/software/jmaplot/current/html/ Wang, Z., Fazio, G. G., Ashby, M. L. N., et al. 2004, ApJS, 154, 193 jmaplot_overview.html Weilbacher, P. M., Streicher, O., Urrutia, T., et al. 2014, in Astronomical

A33, page 11 of 12 A&A 615, A33 (2018)

Data Analysis Software and Systems XXIII, eds. N. Manset, & P. Forshay, Weselak, T., Galazutdinov, G. A., Musaev, F. A., & Krełowski, J. 2004, A&A, ASP Conf. Ser., 485, 451 414, 949 Weilbacher, P. M., Monreal-Ibero, A., Verhamme, A., et al. 2018, A&A, 611, Whitmore, B. C., Chandar, R., Schweizer, F., et al. 2010, AJ, 140, 75 A95 Whitmore, B. C., Brogan, C., Chandar, R., et al. 2014, ApJ, 795, 156 Welty, D. E. 2014, in The Diffuse Interstellar Bands, eds. J. Cami, & N. L. J. Wilson, C. D., Scoville, N., Madden, S. C., & Charmandaris, V. 2000, ApJ, 542, Cox, IAU Symp., 297, 153 120 Welty, D. E., Federman, S. R., Gredel, R., Thorburn, J. A., & Lambert, D. L. Worthey, G., Faber, S. M., Gonzalez, J. J., & Burstein, D. 1994, ApJS, 94, 2006, ApJS, 165, 138 687 Welty, D. E., Ritchey, A. M., Dahlstrom, J. A., & York, D. G. 2014, ApJ, 792, Zaragoza-Cardiel, J., Font, J., Beckman, J. E., et al. 2014, MNRAS, 445, 106 1412 Wendt, M., Husser, T.-O., Kamann, S., et al. 2017, A&A, 607, A133 Zhu, M., Seaquist, E. R., & Kuno, N. 2003, ApJ, 588, 243

A33, page 12 of 12