A newly-discovered young massive star cluster at the far end of the Galactic Bar
We present a near-infrared study of the candidate star cluster Mercer 81, located at the centre of the G338.4+0.1 Hii region, and close to the TeV gamma-ray source HESS 1640-465. Using HST/NICMOS imaging and VLT/ISAAC spectroscopy we have detected a compact and highly extincted cluster of stars, though the bright stars in the centre of the field are in fact foreground objects. The cluster contains nine stars with strong P emission, one of which we identify as a Wolf-Rayet (WR) star, as well as an A-type supergiant. The line-of-sight extinction is very large, , illustrating the challenges of locating young star clusters in the Galactic Plane. From a quantitative analysis of the WR star we argue for a cluster age of 3.7 Myr, and, assuming that all emission-line stars are WRs, a cluster mass of M. A kinematic analysis of the cluster’s surrounding Hii-region shows that the cluster is located in the Galactic disk at a distance of 112 kpc. This places the cluster close to where the far end of the Bar intersects the Norma spiral arm. This cluster, as well as the nearby cluster [DBS2003]179, represent the first detections of active star cluster formation at this side of the Bar, in contrast to the near side which is well known to have recently undergone a M starburst episode.
keywords:(Galaxy:) open clusters and associations: general (Galaxy:) open clusters and associations: individual: Mercer 81 stars: Wolf-Rayet (ISM:) H ii regions ISM: clouds
Young massive star clusters (YMCs – ages 50Myr, masses M) have relevance to many areas of astrophysics. They contain large numbers of massive stars, whose high temperatures, high luminosities, dense winds, and supernova explosions make YMCs a considerable source of mechanical energy, ionizing radiation and chemically processed ejecta. The effect that they have on their surroundings is profound, clearing away the remains of their natal molecular cloud, whilst revealing and triggering subsequent generations of star formation (e.g. González Delgado & Pérez, 2000; Smith, 2006; Davies et al., 2011a; De Marchi et al., 2011). Their large populations of massive stars make them ideal natural laboratories with which to study the evolution of massive stars up to supernova and beyond (e.g. Martins et al., 2007, 2008; Bibby et al., 2008; Davies et al., 2008, 2009). Finally, they can dominate the radiative output of their host galaxies, either through direct ultraviolet and optical emission, or through reprocessed emission in the form of ionised gas or heated dust (Alonso-Herrero et al., 2002).
Our knowledge of our own Galaxy’s population of YMCs is extremely incomplete, in stark contrast to external galaxies (e.g. Bastian et al., 2005; Konstantopoulos et al., 2009). The high levels of interstellar extinction in the plane of the Galaxy have meant that until recently very few clusters were known beyond a distance of 1kpc. Most known clusters beyond this distance were found by targeted searches of, for example, the Galactic Centre (Cotera et al., 1996; Figer et al., 1999), giant Hii-regions (Blum et al., 1999, 2000, 2001), fields around newly-born neutron stars (Fuchs et al., 1999; Vrba et al., 2000), or simply because the foreground extinction was low enough to detect the cluster at optical wavelengths (Westerlund, 1987). However, recent infrared (IR) surveys of the Galactic plane, beginning with 2MASS (Skrutskie et al., 2006), and more recently Spitzer/GLIMPSE (Benjamin et al., 2003) and VVV (Minniti et al., 2010), are at last helping us to uncover the cluster population of the Galactic disk, and affording the opportunity to search the Galactic Plane for clusters in a systematic way.
By-eye and algorithmic searches of survey data (e.g. Ivanov et al., 2002; Dutra et al., 2003; Mercer et al., 2005; Froebrich et al., 2007; Borissova et al., 2011) have yielded over 1,000 candidates for newly discovered star clusters. However, such catalogues inevitably contain large numbers of false positives, due to chance alignments of stars, patchy foreground extinction, and spatially extended emission incorrectly classified as unresolved star clusters. Therefore, these catalogues of candidates must be analysed carefully using multiwavelength survey data and, ultimately, follow-up spectroscopic observations before their distances and physical properties may be derived (e.g. Kurtev et al., 2007; Messineo et al., 2009; Hanson et al., 2010). Only then can they be placed in the framework of the Galaxy’s recent star-forming history.
One such candidate is object #81 from the catalogue of Mercer et al. (2005), known hereafter as Mc81111Objects in this catalogue are referred to by some authors with the prefix GLIMPSE, e.g. GLIMPSE81. This object was found in an algorithmic search of the GLIMPSE survey, by looking for spatial groups of stars with similar photometric properties. The clues to its nature, however, come from cross-correlation with other data in the literature. The object appears to be at the centre of the Hii-region G338.4+0.1, a bubble of warm dust and ionized gas, visible in the SUMSS 843MHz (Bock et al., 1999) and MSX 8m (Price et al., 2001) images, and in more detail in the GLIMPSE 8m image (Fig. 1). The object’s location is close to a supernova remnant (SNR) by Green (2004) from the ‘shell’ morphology of continuum radio emission, but with no spectral index measurement the object could also be a wind-blown bubble.
Close to the centre of the SNR is a high-energy TeV gamma-ray source, HESS J1640-465 (Aharonian et al., 2005), at a distance of 8 (a linear distance of 22pc, if the complex is at a distance of 11kpc – see Sect. 3.3). This source, as with many other TeV sources, is thought to be associated with a pulsar wind nebula (Funk et al., 2007), indicative of recent SN activity. Indeed, TeV emission can be a useful tracer of massive star formation in the Galactic Plane as it is unaffected by absorption, and there are other young massive star clusters in the literature that are known to be associated with such sources – RSGC1 (Figer et al., 2006; Davies et al., 2008), Cl 1813-178 (Messineo et al., 2008), and Westerlund 2 (Aharonian et al., 2007). Finally, in a follow-up of HESS J1640-465, a number of hard X-ray sources were detected in the field of Mc81, with one source being perfectly aligned with the cluster (Landi et al., 2006, see right panel of Fig. 1). Such emission could be explained by either a recently-formed neutron star or a colliding wind binary, with both explanations being indicators of youth (for an analysis of the X-ray emission from another young star cluster, Westerlund 1, see Clark et al., 2008).
Based on this evidence we have extensively followed-up Mc81 with near-infrared (NIR) photometry and spectroscopy, with a view to confirming that the object is indeed a young star cluster, and ultimately to determine the cluster’s physical properties.
2 Observations & data reduction
Images were obtained with HST/NICMOS on 22 October 2008, as part of observing programme #11545 (PI: B. Davies). We used the NIC3 camera which has a field-of-view of 51.251.2 and a pixel scale of 0.2. We observed the cluster through filters F160W and F222M, as well as the narrow-band filters F187N and F190N which are centred on P and the neighbouring continuum respectively. In addition to the cluster we observed a nearby control field through the F160W and F222M filters in order to characterize the foreground population. The fields of observation are indicated in Fig. 1.
Our observations used a spiral dither pattern with six pointings with offset distance was set to 5.07. This sub-pixel dithering technique was designed to minimise the impact of non-uniform intra-pixel sensitivity on our photometry. The MULTIACCUM read modes were used, with the sampling sequences and total integration times that are listed in Table 1.
Our reduction procedure followed the guidelines of the NICMOS Data Handbook v7.0. The standard reduction steps of bias subtraction, dark-current correction and flat-fielding were performed using calnica. Before mosaicing, each dithered observation was subsampled onto a 3 finer grid using bi-linear intepolation to account for the sub-pixel dithering.
Photometry was extracted from the images using the starfinder package which run within IDL (Diolaiti et al., 2000), in conjunction with point-spread functions (PSFs) which were computed for each filter using tinytim. starfinder uses these PSFs to locate stars within each image, and we employed two iteration cycles to fine-tune the astrometry and photometry. Since our fields of observation are not very crowded, we did not use the deblending algorithm, as this was found to produce many false detections. Uncertainties and completeness limits were determined by inserting fake stars into each image and measuring the recovery rate.
Spectroscopic data were taken on the nights of 2009-4-11 and 2009-5-4 as part of ESO observing programme 083.D-0765(A) (PI: E. Puga), using the ISAAC spectrograph on the VLT (Moorwood et al., 1998) in ‘medium’ resolution mode. Our targets were the stars labelled ‘1’ and ‘3’ in Fig. 2, which were determined to be likely cluster members based on their photometric properties (see later). We used the 0.8 slit at three different central wavlengths: 1.71m, 2.09m, and 2.21m, providing a spectral resolution of 4,000. The DITNDITNINT combination for each wavelength setting was (8830s), (8832s), (16824s) respectively. The observations were taken in a ABBA pattern to isolate and subtract sky emission features. In addition to the target stars we observed the B9 v stars Hip090248 and Hip091286 as measures of the telluric absorption, as well as the usual observations of flat-fields, dark frames and arcs for wavelength calibration. To characterize the spatial distortion, a bright field star was stepped along the slit and re-observed multiple times.
The data reduction procedure began with subtraction of nod pairs to remove sky emission, and division by a normalized flat-field. The 2-D frames were then rectified onto an orthogonal grid, using the stepped-star and arc frames to characterize the distortion in the spatial and dispersion directions respectively. This process also wavelength calibrates the data, with r.m.s. residuals which were found to be less than a tenth of a resolution element (10 km s). After rectification, the spectra were extracted and combined.
The telluric standard spectra had their intrinsic H i absoption removed by fitting the lines with Voigt profiles. The telluric spectra were then cross-correlated with the target spectra in the region of isolated telluric features to correct for any residual sub-pixel shifts, before the target spectra were divided through by the telluric spectra.
3 Results & Analysis
The NICMOS images of Mc81 can be seen in Fig. 2. The left panel shows the F222M image of the cluster. In the right panel, we show the difference image [F187N-F190N], which clearly highlights 9 stars with significant P emission222It is possible that He ii emission contributes to the flux in this band, especially if the stars are WRs.. This strongly suggests that these are hot stars with strong winds, and are therefore likely to have ages 10Myr.
In Fig. 3 we show a 3-colour RGB image of F222M (red), F160W (green), and [F187N-F190N] (blue). Around the coordinate centre, there is a clear group of highly reddened stars which form the putative cluster. In this colour scheme, the emission line stars, which are also heavily reddened, show up as pink/magenta. From these data there is already strong evidence for a young, highly reddened cluster of stars in the field of Mc81. The four bright stars to the south of the cluster have a green/yellow colour, indicating that they are in the foreground. Ironically, it is likely that these four stars in part triggered the cluster detection algorithm used by Mercer et al. (2005). These authors claim that they detected an association of 65 stars, though the positions of these stars are not listed, so it is not possible for us to check whether the cluster detected by Mercer et al.’s algorithm bears any relation to the cluster we describe here. For the sake of clarity and consistency, we continue to refer to the cluster studied in this work as Mc81.
The results of the NICMOS photometry are shown in Fig. 4. The left panel shows a colour-magnitude diagram of the stars within 18 of the cluster centre, which we define as the location of the brightest emission-line star. Also shown on the plot are data from an area in the control field of identical size. There is an obvious difference between the two regions, with the ‘cluster’ field showing an excess of stars at ()2.3. The centre panel shows the P excess stars. Those stars with () colours 0.3 and 13 are defined as ‘P emitters’, and are indicated on the plot. The locations and photometry of these stars are listed in Table 2.
Finally, the right-hand panel of Fig. 4 shows the same as the left, after the cluster field has been decontaminated of field stars using the data from the control-field. We do this by eliminating stars from the cluster field which have a control-field star close by in colour-magnitude space. We set this limiting distance to be the larger of either the cluster field star’s 1- photometric errors, or 0.15mags in colour and 0.1mags in magnitude. We also set the restriction that no control-field star can eliminate more than one cluster field star. After decontamination, the cluster sequence at ()2.3 becomes clearer, though there is still some scatter.
|1||16 40 29.83 -46 23 33.9||11.14||8.90||9.67||9.75|
|2||16 40 29.65 -46 23 29.1||13.25||10.59||10.69||11.63|
|3||16 40 30.08 -46 23 11.4||12.97||10.82||10.75||11.60|
|4||16 40 29.65 -46 23 28.7||13.44||10.95||11.15||11.91|
|5||16 40 28.35 -46 23 25.6||14.13||11.42||11.67||12.45|
|6||16 40 29.32 -46 23 11.6||13.46||11.42||11.56||12.14|
|7||16 40 29.60 -46 23 25.6||13.55||11.45||11.56||12.31|
|8||16 40 28.94 -46 23 27.1||14.27||11.46||11.95||12.65|
|9||16 40 29.32 -46 23 38.2||13.76||11.82||12.16||12.54|
|10||16 40 30.05 -46 23 24.3||15.08||12.78||13.41||13.70|
We now present the results of the spectroscopic observations. To classify the spectra, we refer to the spectral atlases of Hanson et al. (2005), Figer et al. (1997), as well as Crowther et al. (2006).
The spectra of the two brightest stars in Mc81 are shown in Fig. 5. The brighter star, Mc81-1, has relatively weak spectral features. The absorption lines of the Hydrogen Brackett series are seen, as well as some faint Mg ii emission. We see no evidence of He I in either absorption or emission. This absence of He I suggests a spectral type later than B5, while the emission lines of Mg ii are indicative of a substantial stellar wind, more typical of supergiants. The ‘ripples’ seen between 2.05-2.08m are due to poor cancellation of the CO absorption. For now, we assign a loose spectral type to this star of late-B/early-A supergiant. Taking the distance and extinction derived in Section 3.3, as well as the bolometric corrections tabulated by Blum et al. (2000) for spectral types B7-A2, we estimate the luminosity of the star to be in the range 5.4–5.8, placing the star close to the empirical stellar luminosity limit at 5.9 (Humphreys & Davidson, 1979). Such stars are also seen in Westerlund 1 (Clark et al., 2005). The star’s proximity to the Humphreys-Davidson limit suggests that the star may be in an unstable phase of evolution, such as a Luminous Blue Variable or Yellow Hypergiant phase, though further spectroscopic and photometric monitoring would be required to verify this.
In contrast to Mc81-1, Mc81-3 has a spectrum rich in strong, broad emission lines. These lines can be attributed to H i, He i, He ii and N iii. The ratio of the He ii 2.189m to the complex at 2.115m, as well as the absorption of He i 2.189m, allow us to tightly constrain the spectral type of this star to be WN7-8.
3.3 Extinction and distance
To calculate the extinction, we first determine the reddening of the cluster sequence from the right-hand panel of Fig. 4. The average colour of the stars in the decontaminated cluster field is ()=2.30.3. If we make the approximation that all main-sequence stars that we detect should have colours of approximately zero, the observed average colour is due to extinction. We can then determine the extinction from the following relation,
with 1.60m and 2.22m, i.e. the wavelengths of the NICMOS F160W and F222M filters. The parameter has been studied by numerous authors in recent years (see e.g. Stead & Hoare, 2009, and references therein), with the most contemporary measurements converging on . This therefore implies that the extinction towards Mc81 is . Extrapolating this extinction to the optical is known to be highly uncertain, but we estimate 333Using the value of from Rieke & Lebofsky (1985), we find , and . Mc81 is therefore one of the most heavily reddened clusters known, with an extinction comparable to that of the Galactic Centre.
We estimate the distance to the cluster from the radial velocity of the surrounding molecular cloud. Caswell & Haynes (1987) studied the radio recombination line emission of the two clouds of ionized gas either side of the cluster, G338.4+0.2 and G338.4+0.1 (see Fig. 1), finding velocities relative to the local standard of rest of =-29 km s and -37 km s respectively. In addition, a number of massive young stellar objects (YSOs) and compact Hii-regions have been detected in the region by the Red MSX Source (RMS) Survey (Hoare et al., 2005; Urquhart et al., 2007a, b), with an average radial velocity of = -35.12.8 km s.
Comparing this average to the Galactic rotation curve of Brand & Blitz (1993), using a Galacto-centric distance of 7.60.3 kpc and a rotational velocity of 2147 km s (Kothes & Dougherty, 2007, and references therein), we find near and far kinematic distances of 3.8kpc and 11kpc. To resolve the near/far ambiguity, we refer to the Southern Galactic Plane Survey (SGPS) of Hi. These data were recently analysed by Lemiere et al. (2009), who found that for the two large Hii-regions G338.4+0.2 and G338.4+0.1, Hi absorption components could be seen at several velocities up to the tangent point velocity of 130 km s. This indicates the Hii-regions, and by association the YSOs and the central star cluster, lay beyond the tangent point. From this we conclude that the clusters lay at the far-side distance of 11 kpc. At this distance, the uncertainty is dominated by the systematic uncertainties in the Galactic rotation curve, which by its nature is difficult to quantify. However, if we assume that the system may have a peculiar velocity of up to 20 km s (Russeil, 2003), this gives an uncertainty on this distance of 2 kpc.
We can check this distance by calculating the absolute brightness of the WNL star and comparing to similar objects. Using the extinction calculated in the previous section and a distance of 112 kpc, we find an absolute magnitude for Mc81-2 of . By comparison, Galactic WNL stars are typically found to have (Crowther et al., 2006). Mc81-2 is therefore somewhat luminous for its spectral type, though it is within the errors for other Galactic WNL stars.
3.4 Cluster size
In Fig. 6, we illustrate the physical extent of the cluster. The figure shows the locations of all stars in the NICMOS field-of-view which are brighter than the 50% completeness limit (), overlayed with those stars which have colours consistent with the cluster (i.e. ). The figure shows that there is a clear overdensity of stars at the coordinate centre (defined as the position of star Mc81-2).
To measure the size and morphology of this overdensity, we first made a map of the stellar density by dividing the field up into square bins of size 33 and counting the number of stars per bin. To reduce noise, this map was then smoothed, such that the effective resolution of the map was 6 (illustrated in the bottom corner of Fig. 6). This resolution size was chosen as a trade-off between spatial resolution and signal strength, though our results were robust to changes in this parameter. The ambient stellar density was found by computing the background level of this map using the GSFC IDL routine sky. Finally, we computed isodensity contours in this map at percentiles of the maximum stellar density.
We defined the size of the cluster where the stellar density drops to 50% of its maximum value, which we deem to be roughly equivalant to its half-light radius444Ideally, to measure the half-light radius one would measure the cumulative surface brightness out to a distance where it becomes asymptotic. However, the field-of-view of our observations is too small to do this.. Once deconvolved with the effective spatial resolution, we find that the cluster has major and minor axes of 29 18. At a distance of 11kpc, this corresponds to 1.51.0 pc, and so is comparable to other young Galactic clusters which typically have diameters between 1-2pc (e.g. Trumpler 14, Westerlund 1; Figer, 2008).
|( km s)||1350 100|
3.5 Quantitative spectral analysis
To model the WNL star in Mercer81 and estimate its physical parameters, we proceed as in Najarro et al. (2004, 2009). Briefly, we have used CMFGEN, the iterative, non-LTE line blanketing method presented by Hillier & Miller (1998) which solves the radiative transfer equation in the co-moving frame and in spherical geometry for the expanding atmospheres of early-type stars. The model is prescribed by the stellar radius, , the stellar luminosity, , the mass-loss rate, , the velocity field, (defined by the terminal wind speed and the wind acceleration parameter ), the volume filling factor characterizing the clumping of the stellar wind, f(r), and elemental abundances. Hillier & Miller (1998, 1999) present a detailed discussion of the code. For the present analysis, we have assumed the atmosphere to be composed of H, He, C, N, O, Si, S, Fe and Ni. Observational constraints are provided by the H, K-band spectra of the stars and the dereddened F166W, F190N and F222M magnitudes.
Given the extreme sensitivity of the and -Band He i and He ii line profiles ratios in this parameter domain to changes in temperature, we estimate our errors in the temperature to be below 1000 K. Likewise, the relative strengths between the H and He lines constrain the H/He ratio to be within 0.5 and 1.0 by mass. The error on and hence on and is dominated by those in the assumed distance and the slope of the extinction law.
The best-fitting model is overplotted in Fig. 5. The model provides a good fit to the features of the observed spectrum, with the exception of the Br10 line, which is blended with emission from N iv/C iv/O iv. This discrepancy is due primarily to the deficiencies in the CNO iv model atoms, the correction of which is beyond the scope of the current work. The model’s physical parameters are given in Table 3, and are typical for a late-type WN star. For completeness, we list the temperature at an optical depth of , which is comparable to the star’s effective temperature; and the temperature at which is comparable to the hydrostatic temperature of stellar structure models. In the following Section we use these results to estimate the age of Mc81-3, and therefore of the cluster itself.
3.6 Cluster age
In order to contrain the age of the cluster, we make a quantitative comparison between the physical properties of the WNL star Mc81-3 and the predictions of stellar evolutionary models. Under the assumption that the star and the host cluster are coeval, we can then estimate the cluster’s age.
For this analysis, the models we have chosen are those of Meynet & Maeder (2000) which are optimized for massive stars. In our method, we linearly interpolate these mass tracks at intervals of 1M and yrs. For each point on each interpolated mass track, we then calculate the probability that there is a match between the mass track and Mc81-3, based on the star’s luminosity , temperature , and H/He ratio derived in the previous Section. For the star’s temperature, we use the temperature at an optical depth of , since this is more comparable to the hydrostatic temperature calculated by Meynet & Maeder (2000).
Figure 7 shows the inter-related behaviour of the three variables , and H/He for models with a range of initial stellar masses. We also plot the derived physical parameters of Mc81-3. The plot shows that, although the star’s luminosity and H/He ratio place it on the 120M mass track, the temperature of the star (illustrated by the colour of the plotting symbol) more closely matches the 60M track.
We assume that the errors on , and are gaussian, and therefore the probability of a match between Mc81-3 and the mass track of initial mass at time is given by,
where is the observed quantity (either , or ), is its associated uncertainty, and is the corresponding quantity predicted by the model mass track. Each term is therefore weighted by its associated uncertainty.
In Fig. 8 we plot how the probability varies across the 2-D plane of mass and stellar age for the rotating models of Meynet & Maeder (2000). The maximum probability () is obtained for an initial mass of 62M and an age of 3.7Myr. The morphology of the iso-probability contours are highly non-gaussian, so for the experimental uncertainty we cannot simply compute the standard deviation. Instead, we take the iso-probability contour at 50% and determine the minimum and maximum values of mass and age for that contour. In this way, for Mc81-3 we find M and an age of Myr. Using the non-rotating versions of the same stellar evolution models produces a slightly different morphology to the probability distribution, with a reduced probability, but with a best-fitting age and mass that do not differ significantly from that derived using the rotating models.
The value we obtain for Mc81-3’s age can be understood through a simple qualitative analysis of the star’s parameters. The high luminosity clearly favours high initial masses, and therefore a young age. In addition, the He enrichment indicates an object which is in an advanced evolutionary state, and so older than 2Myr, but younger than the total lifetime of a high-mass star, and so therefore younger than 5Myr.
3.6.1 The impact of binary evolution on our derived cluster age
Throughout this analysis we have assumed that Mc81-3 has evolved as a single star. However, there is the possibility that the star is in an interacting binary: both Landi et al. (2006) and Funk et al. (2007) have detected X-ray emission from the centre of the cluster, which could be evidence of a colliding wind binary system. As shown by Eldridge (2009) for the case of Vel, including the effects of binary evolution in the analysis can alter the derived age of a star.
In the case of Mc81-3, the star’s high luminosity places a strong constraint on the initial mass of the star, and hence on the upper limit of its age. Though binary evolution can affect the surface abundances and temperature of a star, and prolong its lifetime, it is unlikely to increase the star’s maximum luminosity by more than 0.1dex, which is governed primarily by the initial stellar mass (all other parameters being equal) (Eldridge et al., 2008). An exception to this would be if two stars merged to produce a completely rejeuvenated and more massive star. In the absence of any evidence for such an event in the history of Mc81-3, we maintain that the upper limit to the star’s age is that derived in the previous paragraph.
The lower limit to the cluster age may be reduced if Mc81-3 is in an interacting binary system. Mass transfer from the primary to the secondary star may speed up the rate at which H is depleted from the primary’s surface. In this case, we would underestimate the stellar (and hence cluster) age by using single star evolution models in our analysis.
However, a simple morphological analysis of the nebula surrounding Mc81, and a comparison to similar systems, serves as a sanity check on our age estimate. The cluster is located at the centre of a cavity, which was presumably evacuated by the winds, ionizing radiation and SNe explosions of the most massive stars in the cluster. At the periphery of the cavity evidence of further generations of star formation is seen, which may or may not have been triggered by feedback from the cluster. This morphology is reminiscent of other cluster + nebula systems such as G305, NGC 3603 and NGC 346 to name but a few. Ages of these other systems are commonly found to be 2-4Myr (Davies et al., 2011a; Harayama et al., 2008; Bouret et al., 2003), and therefore are consistent with our estimate for Mc81.
3.7 Cluster mass
For the cluster mass, it is difficult to make an accurate estimate without further spectroscopy of the stars in the cluster. We can however make a rough estimate of the cluster mass from the emission line stars. If we assume that all the nine strong P emitters listed in Table 2 are WRs, then since the age we derive for Mc81 is roughly the same as that of Westerlund 1 (Wd1) (3-5Myr, Crowther et al., 2006; Brandner et al., 2008) which has 27 WRs, this suggests that Mc81 may be a factor of 3 less massive than Wd1. As most estimates of Wd1’s mass are around M (Clark et al., 2005; Brandner et al., 2008) this implies that the mass of Mc81 is a few M. We stress however that this is only an order-of-magnitude estimate; a more precise measurement of the cluster mass awaits further analysis of its stellar population.
|Danks 1 & 2||305.0||4.0||9.5||23|
|DBS2003179||347.6||9.0||15.2||26, This work.|
References: 1: Bibby et al. (2008); 2: Blum et al. (2001);
3: Messineo et al. (2008); 4: Hanson et al. (1996); 5: Messineo et al. (2010);
6: Blum et al. (2000); 7: Davies et al. (2008); 8: Davies et al. (2007);
9: Clark et al. (2009b); 10: Negueruela et al. (2010); 11:Negueruela et al. (2011);
12: Messineo et al. (2009); 13: Blum et al. (1999); 14: Davies et al. (2009);
15: Hanson et al. (2010); 16: Hanson (2003); 17: Currie et al. (2010);
18: Zhu et al. (2009); 19: Rauw et al. (2007); 20: Ascenso et al. (2007);
21: Harayama et al. (2008); 22: Kurtev et al. (2007);
23: Davies et al. (2011a); 24: Kothes & Dougherty (2007);
25: Raboud et al. (1997); 26: Borissova et al. (2008).
4.1 Location in the Galaxy
With the many recent discoveries of young star clusters in the Galaxy, we can now begin to build up a picture of the Galaxy’s recent cluster formation. In Fig. 9 we plot the locations of all known young Galactic clusters with distances from the Sun greater than 2 kpc. All clusters in the plot are thought to have masses in excess of M and ages 20Myr. The references for each data-point are listed in Table 4. We have colour-coded each data-point according to its visual extinction, which we have calculated in a homogeneous way from each cluster’s , measured either from the references listed in 4 or from 2MASS photometry, and an extinction law slope of (see Sect. 3.3).
Figure 9 shows that there are now a significant number of reddened star clusters known at Galactic longitudes between 1050. This line-of-sight corresponds to the tangent of the Scutum-Crux arm, as well as the near end of the Galactic Bar (Benjamin et al., 2005). Since one would expect the star formation rate to be comparatively high in this location of the Galaxy, it is also reasonable to expect it to be rich in young star clusters. Indeed, the region hosts five known YMCs, plus a substantial field population of Red Supergiants, indicating a starburst episode of M around 20Myr ago (Garzon et al., 1997; López-Corredoira et al., 1999; Figer et al., 2006; Davies et al., 2007; Clark et al., 2009b; Negueruela et al., 2010, 2011).
However, less is known about the opposite side of the Galactic Centre and far end of the Bar. There are two possible reasons for this: firstly, the larger distance, high extinction and larger number of foreground stars (due to the intervening Galactic Bulge) make it more difficult to pick out clusters in by-eye searches. Indeed, it is unlikely that Mc81 would have been found were it not for the four bright foreground stars (see Sect. 3.1). Secondly, the fact that no star clusters are known in this direction means that investigators are less likely to search this region. This is in contrast to the near end of the Bar, where the initial discovery of the cluster RSGC1 in this region by Figer et al. (2006) led to the subsequent discoveries of a further 4 clusters within the same complex 555Though RSGC1 and RSGC2 were first recognized as associations of stars by Bica et al. (2003) and Stephenson (1990) respectively, the nature of each cluster was not understood until later..
From our distance estimate of Mc81, we can place this cluster close to where we suppose the far-end of the Bar may be, assuming an azimuthal angle of 44° and a bar length of 4.4 kpc (Benjamin et al., 2005). Another cluster nearby, which may too trace the end of the Bar, is [DBS2003]179. The distance to this cluster is not well known, and is based on spectro-photometric distance estimates for stars with unknown luminosity classes (Borissova et al., 2008). We have reassesed the distance to this cluster using a similar methodology that we have presented here for Mc81. Specifically, we assume that the cluster is physically associated to its nearby molecular cloud, and use the massive YSOs detected in the cloud to determine the systemic radial velocity. We then use the SGPS survey (McClure-Griffiths et al., 2005) to measure the velocity spectrum of the intervening H i gas to resolve the distance ambiguity. The average radial velocities of the YSOs (=-363 km s), combined with the H i absorption which is seen up to tangent-point velocities of 130 km s, give a far-side kinematic distance of 9 kpc, again with a 2 km s uncertainty to allow for deviations from the Galactic rotation curve (see Sect. 3.3). This is within the errors of the spectro-photometric distance of 7.9 kpc derived by Borissova et al. (2008).
These two clusters – Mc81 and [DBS2003]179 – are then the first young star clusters to be discovered in this region of the Galaxy. In addition to these clusters a group of giant Hii-regions, of which G338.4+0.1 is one (Russeil, 2003), suggest that this region may be an active star-formation site, similar to the region of the Scutum-Crux tangent at the opposite end of the Bar. As such, future targeted surveys of this region may unearth a number of other such objects.
4.2 Association with HESS 1640-465
As was noted in the introduction, the Mc81 cluster is located only a few arcminutes from the TeV source HESS 1640-465, which is likely to be powered by a neutron star. If the two were associated, it would allow us to estimate the initial mass of the neutron star’s progenitor. Here we discuss the possible association between the two objects.
HESS 1640-465 is known to have a counterpart source at GeV energies (Slane et al., 2010), and in the X-ray, detected by Swift, XMM and Chandra (Landi et al., 2006; Funk et al., 2007; Lemiere et al., 2009). Analysis of the XMM X-ray spectrum yielded a column density of cm or cm, depending on whether a power-law or absorbed black-body model was used (Funk et al., 2007). However, Lemiere et al. (2009) analysed the Chandra data and found that they required a higher column density of cm to fit the data. Assuming a standard calibration between and optical extinction of cm (Predehl & Schmitt, 1995), this implies an visual extinction of between 20 and 70 mags, depending on the model for the X-ray emission. Though the errors are large, this is consistent with our measurment of the extinction to the cluster of .
From this evidence, it seems likely that HESS 1640-465 is associated with the G338.4+0.1 Hii-region surrounding Mc81. However, its connection with the cluster itself is not clear. If the central source of HESS 1640-465 is a neutron star (as seems likely), and the progenitor was born with the cluster but was ejected, there are two possibilities: either the progenitor was dynamically ejected from the cluster; or it received a kick from the supernova (SN). The location of HESS 1640-465 at the centre of SNR 338.3+0.0 (Green, 2004) provides circumstantial evidence against the latter explanation, since this suggests that the progenitor exploded close to its present location. If the progenitor formed with the cluster but was ejected at a time ago, the ejection velocity is km s(assuming a projected distance of 22 pc). Therefore, it is entirely plausible that the progenitor star formed along with the rest of the stars in Mc81 and was dynamically ejected during the formation of the cluster.
Finally, there is the possibility that HESS 1640-465 formed out of the same molecular cloud as Mc81, and at a similar time, but that the two formed independently of one another. The morphology of the G338.4 region suggests inside-out star-formation, with the 3Myr old cluster in the centre and a series of YSOs and UC-Hii regions at the periphery of the surrounding cavity, which have ages of a few yrs (Davies et al., 2011b). The location of HESS 1640-465 does not fit this picture, since the progenitor star must have formed at least 2Myr ago, which is before the first SNe occured in Mc81. However, there are other known instances of ‘multi-seeded’ star formation, where collapse occurs at multiple causally-unrelated sites across the host GMC (e.g. W51, Clark et al., 2009a).
In summary, we conclude that HESS 1640-465 is likely associated with the star-formation region of G338.4+0.1. However, we are unable to make a definitive association with the star cluster Mc81, and so we are unable to use the age of the cluster to estimate the mass of the neutron star’s progenitor, as we were in the cases of e.g. RSGC1 and Cl 1900+14 (Davies et al., 2008, 2009).
We have provided a near-infrared photometric and spectroscopic investigation of the candidate star cluster Mercer 81 (Mc81). We find that that a highly extincted () cluster exists in the field identified by Mercer et al. (2005), but that the bright four stars at the centre of the field are unrelated foreground objects. The cluster is located at the centre of a cavity in a large Hii-region in the direction of G338.4+0.1, with evidence of on-going star formation in the periphery of the cloud. Our analysis of the cluster has revealed nine stars with strong P emission, one of which we identify spectroscopically as a late-type N-rich Wolf-Rayet star (WNL), in addition to a luminous early A-type supergiant. Via detailed modelling of the WNL star’s spectrum we estimate an age for the cluster of Myr. Under the assumption that the stars with strong line emission are WRs, we have made an order-of-magnitude estimate of the cluster’s mass of M.
From a kinematic analysis of the host cloud, we obtain a distance to the host star-forming complex of 112 kpc. Our distance estimate therefore places the G338.4+0.1 complex in the same region of the Galaxy as the far end of the Galactic Bar. The recent detection of another star cluster close to this location, as well as other giant Hii-regions, suggest that this region of the Galaxy may be as active in star-formation as the opposite end of the Bar, where a M starburst event is known to have occurred in the last 20Myr. A targeted search of the far end of the Bar will likely uncover many more young star clusters, though the high extinction, large distance and dense stellar field of the intervening Galactic Bulge will mean that such a search will require considerable observational effort.
We thank the anonymous referee for comments and suggestions which helped us improve the paper. BD is supported by a fellowship from the Royal Astronomical Society. This work is in part based on observations made with the NASA/ESA Hubble Space Telescope, obtained at the Space Telescope Science Institute, which is operated by the Association of Universities for Research in Astronomy, Inc., under NASA contract NAS 5-26555. These observations are associated with program #11545. Support for program #11545 was provided by NASA through a grant from the Space Telescope Science Institute, which is operated by the Association of Universities for Research in Astronomy, Inc., under NASA contract NAS 5-26555. This work is in part based on observations collected at the European Organisation for Astronomical Research in the Southern Hemisphere, Chile, under programme number 083.D-0765(A). Financial support from the Spanish Ministerio de Ciencia e Innovación under projects AYA2008-06166-C03-02 and AYA2010-21697-C05-01 is acknowledged.
- Aharonian et al. (2005) Aharonian, F., Akhperjanian, A. G., Aye, K.-M., et al. 2005, Science, 307, 1938
- Aharonian et al. (2007) Aharonian, F., Akhperjanian, A. G., Bazer-Bachi, A. R., et al. 2007, A&A, 467, 1075
- Alonso-Herrero et al. (2002) Alonso-Herrero, A., Rieke, G. H., Rieke, M. J., & Scoville, N. Z. 2002, AJ, 124, 166
- Ascenso et al. (2007) Ascenso, J., Alves, J., Vicente, S., & Lago, M. T. V. T. 2007, A&A, 476, 199
- Bastian et al. (2005) Bastian, N., Gieles, M., Efremov, Y. N., & Lamers, H. J. G. L. M. 2005, A&A, 443, 79
- Benjamin et al. (2003) Benjamin, R. A., Churchwell, E., Babler, B. L., et al. 2003, PASP, 115, 953
- Benjamin et al. (2005) Benjamin, R. A., Churchwell, E., Babler, B. L., et al. 2005, ApJ, 630, L149
- Bibby et al. (2008) Bibby, J. L., Crowther, P. A., Furness, J. P., & Clark, J. S. 2008, MNRAS, 386, L23
- Bica et al. (2003) Bica, E., Dutra, C. M., Soares, J., & Barbuy, B. 2003, A&A, 404, 223
- Blum et al. (2000) Blum, R. D., Conti, P. S., & Damineli, A. 2000, AJ, 119, 1860
- Blum et al. (1999) Blum, R. D., Damineli, A., & Conti, P. S. 1999, AJ, 117, 1392
- Blum et al. (2001) Blum, R. D., Damineli, A., & Conti, P. S. 2001, AJ, 121, 3149
- Bock et al. (1999) Bock, D. C.-J., Large, M. I., & Sadler, E. M. 1999, AJ, 117, 1578
- Borissova et al. (2011) Borissova, J., Bonatto, C., Kurtev, R., et al. 2011, ArXiv e-prints
- Borissova et al. (2008) Borissova, J., Ivanov, V. D., Hanson, M. M., et al. 2008, A&A, 488, 151
- Bouret et al. (2003) Bouret, J.-C., Lanz, T., Hillier, D. J., et al. 2003, ApJ, 595, 1182
- Brand & Blitz (1993) Brand, J. & Blitz, L. 1993, A&A, 275, 67
- Brandner et al. (2008) Brandner, W., Clark, J. S., Stolte, A., et al. 2008, A&A, 478, 137
- Caswell & Haynes (1987) Caswell, J. L. & Haynes, R. F. 1987, A&A, 171, 261
- Clark et al. (2009a) Clark, J. S., Davies, B., Najarro, F., et al. 2009a, A&A, 504, 429
- Clark et al. (2008) Clark, J. S., Muno, M. P., Negueruela, I., et al. 2008, A&A, 477, 147
- Clark et al. (2005) Clark, J. S., Negueruela, I., Crowther, P. A., & Goodwin, S. P. 2005, A&A, 434, 949
- Clark et al. (2009b) Clark, J. S., Negueruela, I., Davies, B., et al. 2009b, A&A, 498, 109
- Cordes & Lazio (2002) Cordes, J. M. & Lazio, T. J. W. 2002, ArXiv Astrophysics e-prints
- Cotera et al. (1996) Cotera, A. S., Erickson, E. F., Colgan, S. W. J., et al. 1996, ApJ, 461, 750
- Crowther et al. (2006) Crowther, P. A., Hadfield, L. J., Clark, J. S., Negueruela, I., & Vacca, W. D. 2006, MNRAS, 372, 1407
- Currie et al. (2010) Currie, T., Hernandez, J., Irwin, J., et al. 2010, ApJS, 186, 191
- Davies et al. (2011a) Davies, B., Clark, J. S., Trombley, C., et al. 2011a, ArXiv e-prints
- Davies et al. (2009) Davies, B., Figer, D. F., Kudritzki, R., et al. 2009, ApJ, 707, 844
- Davies et al. (2007) Davies, B., Figer, D. F., Kudritzki, R.-P., et al. 2007, ApJ, 671, 781
- Davies et al. (2008) Davies, B., Figer, D. F., Law, C. J., et al. 2008, ApJ, 676, 1016
- Davies et al. (2011b) Davies, B., Hoare, M. G., Lumsden, S. L., et al. 2011b, MNRAS, 416, 972
- De Marchi et al. (2011) De Marchi, G., Paresce, F., Panagia, N., et al. 2011, ApJ, 739, 27
- Diolaiti et al. (2000) Diolaiti, E., Bendinelli, O., Bonaccini, D., et al. 2000, A&AS, 147, 335
- Dutra et al. (2003) Dutra, C. M., Bica, E., Soares, J., & Barbuy, B. 2003, A&A, 400, 533
- Eldridge (2009) Eldridge, J. J. 2009, MNRAS, 400, L20
- Eldridge et al. (2008) Eldridge, J. J., Izzard, R. G., & Tout, C. A. 2008, MNRAS, 384, 1109
- Figer (2008) Figer, D. F. 2008, in IAU Symposium, Vol. 250, IAU Symposium, ed. F. Bresolin, P. A. Crowther, & J. Puls, 247–256
- Figer et al. (1999) Figer, D. F., Kim, S. S., Morris, M., et al. 1999, ApJ, 525, 750
- Figer et al. (2006) Figer, D. F., MacKenty, J. W., Robberto, M., et al. 2006, ApJ, 643, 1166
- Figer et al. (1997) Figer, D. F., McLean, I. S., & Najarro, F. 1997, ApJ, 486, 420
- Froebrich et al. (2007) Froebrich, D., Scholz, A., & Raftery, C. L. 2007, MNRAS, 374, 399
- Fuchs et al. (1999) Fuchs, Y., Mirabel, F., Chaty, S., et al. 1999, A&A, 350, 891
- Funk et al. (2007) Funk, S., Hinton, J. A., Pühlhofer, G., et al. 2007, ApJ, 662, 517
- Garzon et al. (1997) Garzon, F., Lopez-Corredoira, M., Hammersley, P., et al. 1997, ApJ, 491, L31+
- González Delgado & Pérez (2000) González Delgado, R. M. & Pérez, E. 2000, MNRAS, 317, 64
- Green (2004) Green, D. A. 2004, Bulletin of the Astronomical Society of India, 32, 335
- Hanson (2003) Hanson, M. M. 2003, ApJ, 597, 957
- Hanson et al. (1996) Hanson, M. M., Conti, P. S., & Rieke, M. J. 1996, ApJS, 107, 281
- Hanson et al. (2005) Hanson, M. M., Kudritzki, R.-P., Kenworthy, M. A., Puls, J., & Tokunaga, A. T. 2005, ApJS, 161, 154
- Hanson et al. (2010) Hanson, M. M., Kurtev, R., Borissova, J., et al. 2010, A&A, 516, A35+
- Harayama et al. (2008) Harayama, Y., Eisenhauer, F., & Martins, F. 2008, ApJ, 675, 1319
- Hillier & Miller (1998) Hillier, D. J. & Miller, D. L. 1998, ApJ, 496, 407
- Hillier & Miller (1999) Hillier, D. J. & Miller, D. L. 1999, ApJ, 519, 354
- Hoare et al. (2005) Hoare, M. G., Lumsden, S. L., Oudmaijer, R. D., et al. 2005, in IAU Symposium, Vol. 227, Massive Star Birth: A Crossroads of Astrophysics, ed. R. Cesaroni, M. Felli, E. Churchwell, & M. Walmsley, 370–375
- Humphreys & Davidson (1979) Humphreys, R. M. & Davidson, K. 1979, ApJ, 232, 409
- Ivanov et al. (2002) Ivanov, V. D., Borissova, J., Pessev, P., Ivanov, G. R., & Kurtev, R. 2002, A&A, 394, L1
- Konstantopoulos et al. (2009) Konstantopoulos, I. S., Bastian, N., Smith, L. J., et al. 2009, ApJ, 701, 1015
- Kothes & Dougherty (2007) Kothes, R. & Dougherty, S. M. 2007, A&A, 468, 993
- Kurtev et al. (2007) Kurtev, R., Borissova, J., Georgiev, L., Ortolani, S., & Ivanov, V. D. 2007, A&A, 475, 209
- Landi et al. (2006) Landi, R., Bassani, L., Malizia, A., et al. 2006, ApJ, 651, 190
- Lemiere et al. (2009) Lemiere, A., Slane, P., Gaensler, B. M., & Murray, S. 2009, ApJ, 706, 1269
- López-Corredoira et al. (1999) López-Corredoira, M., Garzón, F., Beckman, J. E., et al. 1999, AJ, 118, 381
- Martins et al. (2007) Martins, F., Genzel, R., Hillier, D. J., et al. 2007, A&A, 468, 233
- Martins et al. (2008) Martins, F., Hillier, D. J., Paumard, T., et al. 2008, A&A, 478, 219
- McClure-Griffiths et al. (2005) McClure-Griffiths, N. M., Dickey, J. M., Gaensler, B. M., et al. 2005, ApJS, 158, 178
- Mercer et al. (2005) Mercer, E. P., Clemens, D. P., Meade, M. R., et al. 2005, ApJ, 635, 560
- Messineo et al. (2009) Messineo, M., Davies, B., Ivanov, V. D., et al. 2009, ApJ, 697, 701
- Messineo et al. (2010) Messineo, M., Figer, D. F., Davies, B., et al. 2010, ApJ, 708, 1241
- Messineo et al. (2008) Messineo, M., Figer, D. F., Davies, B., et al. 2008, ApJ, 683, L155
- Meynet & Maeder (2000) Meynet, G. & Maeder, A. 2000, A&A, 361, 101
- Minniti et al. (2010) Minniti, D., Lucas, P. W., Emerson, J. P., et al. 2010, New A, 15, 433
- Moorwood et al. (1998) Moorwood, A., Cuby, J., Biereichel, P., et al. 1998, The Messenger, 94, 7
- Najarro et al. (2009) Najarro, F., Figer, D. F., Hillier, D. J., Geballe, T. R., & Kudritzki, R. P. 2009, ApJ, 691, 1816
- Najarro et al. (2004) Najarro, F., Figer, D. F., Hillier, D. J., & Kudritzki, R. P. 2004, ApJ, 611, L105
- Negueruela et al. (2011) Negueruela, I., González-Fernández, C., Marco, A., & Clark, J. S. 2011, A&A, 528, A59+
- Negueruela et al. (2010) Negueruela, I., González-Fernández, C., Marco, A., Clark, J. S., & Martínez-Núñez, S. 2010, A&A, 513, A74+
- Predehl & Schmitt (1995) Predehl, P. & Schmitt, J. H. M. M. 1995, A&A, 293, 889
- Price et al. (2001) Price, S. D., Egan, M. P., Carey, S. J., Mizuno, D. R., & Kuchar, T. A. 2001, AJ, 121, 2819
- Raboud et al. (1997) Raboud, D., Cramer, N., & Bernasconi, P. A. 1997, A&A, 325, 167
- Rauw et al. (2007) Rauw, G., Manfroid, J., Gosset, E., et al. 2007, A&A, 463, 981
- Rieke & Lebofsky (1985) Rieke, G. H. & Lebofsky, M. J. 1985, ApJ, 288, 618
- Russeil (2003) Russeil, D. 2003, A&A, 397, 133
- Skrutskie et al. (2006) Skrutskie, M. F., Cutri, R. M., Stiening, R., et al. 2006, AJ, 131, 1163
- Slane et al. (2010) Slane, P., Castro, D., Funk, S., et al. 2010, ApJ, 720, 266
- Smith (2006) Smith, N. 2006, MNRAS, 367, 763
- Stead & Hoare (2009) Stead, J. J. & Hoare, M. G. 2009, MNRAS, 400, 731
- Stephenson (1990) Stephenson, C. B. 1990, AJ, 99, 1867
- Urquhart et al. (2007a) Urquhart, J. S., Busfield, A. L., Hoare, M. G., et al. 2007a, A&A, 461, 11
- Urquhart et al. (2007b) Urquhart, J. S., Busfield, A. L., Hoare, M. G., et al. 2007b, A&A, 474, 891
- Vrba et al. (2000) Vrba, F. J., Henden, A. A., Luginbuhl, C. B., et al. 2000, ApJ, 533, L17
- Westerlund (1987) Westerlund, B. E. 1987, A&AS, 70, 311
- Zhu et al. (2009) Zhu, Q., Davies, B., Figer, D. F., & Trombley, C. 2009, ApJ, 702, 929