EDP Sciences logo
Subscriber Authentication Point
Search
Menu
Home All issues Volume 639 (July 2020) A&A, 639 (2020) A65 Full HTML
Free Access
Issue
A&A
Volume 639, July 2020
Article Number A65
Number of page(s) 23
Section Interstellar and circumstellar matter
DOI https://doi.org/10.1051/0004-6361/202037817
Published online 09 July 2020

© ESO 2020

1 Introduction

Some of the Galactic interstellar clouds are so dense that when they are seen against the Galactic mid-infrared (mid-IR) background radiation field, they manifest themselves as dark absorption features (Pérault et al. 1996; Egan et al. 1998; Simon et al. 2006; Peretto & Fuller 2009). These clouds are known as infrared dark clouds (IRDCs).

Ever since the IRDCs were uncovered almost a quarter-century ago, we have seen a growing interest in learning more about them. Studies of IRDCs have addressed a wide range of their properties, all the way from their formation (e.g. Jiménez-Serra et al. 2010; Henshaw et al. 2013), Galactic distribution (Jackson et al. 2008; Finn et al. 2013; Giannetti et al. 2015), magnetic fields (e.g. Hoq et al. 2017; Soam et al. 2019), fragmentation (e.g. Jackson et al. 2010; Wang et al. 2011; Busquet et al. 2016; Henshaw et al. 2016; Tang et al. 2019), substructure properties (e.g. Rathborne et al. 2010; Ragan et al. 2013), star formation (e.g. Rathborne et al. 2005, 2006; Chambers et al. 2009; Kauffmann & Pillai 2010), maser emission (e.g. Wang et al. 2006), and dust properties (e.g. Juvela et al. 2018) up to their chemistry (e.g. Vasyunina et al. 2011, 2014; Sanhueza et al. 2012; Gerner et al. 2014; Zeng et al. 2017; Colzi et al. 2018; Taniguchi et al. 2019).

Observations of IRDCs have shown that they often appear to be filamentary structures, which could be the results of converging atomic gas flows or collisions between precursor interstellar clouds that led to the formation of the IRDC filament (e.g. Jiménez-Serra et al. 2010; cf. Heitsch et al. 2006). Moreover, the IRDC filaments appear to be fragmented along their longest axis, potentially via the so-called sausage-type fluid instability in cylindrical geometry, and the substructures thus formed, known as clumps, are found to be hierarchically fragmented down to smaller units, the dense cores (e.g. Jackson et al. 2010; Wang et al. 2011; Beuther et al. 2015; Mattern et al. 2018; Svoboda et al. 2019). Some of the dense substructures of IRDCs are found to be associated with early stages of high-mass star formation (e.g. Rathborne et al. 2006; Beuther & Steinacker 2007; Chambers et al. 2009; Battersby et al. 2010). This makes IRDCs very useful targets for the star formation research because the process(es) of high-mass star formation still remains inconclusive (e.g. Motte et al. 2018 for a review; Padoan et al. 2019).

In this paper, we present a study of the chemical properties of the filamentary IRDC G304.74+01.32. This IRDC has been the target of submillimetre and millimetre dust continuum studies (Beltrán et al. 2006; Miettinen & Harju 2010; Miettinen 2018), and molecular spectral line observations (Miettinen 2012). Most recently, Miettinen (2018) found that 36% of the clumps along G304.74+01.32 are fragmented into smaller cores, and that 65% of those cores are associated with IR sources detected with the Wide-field Infrared Survey Explorer (WISE; Wright et al. 2010). Those results strongly point to a hierarchical fragmentation of the filament. Miettinen (2018) also found that G304.74+01.32 has a seahorse-like morphology in the Herschel1 (Pilbratt et al. 2010) far-IR and submillimetre images (Fig. 1; cf. Fig. 2 in Miettinen 2018) and, therefore, proposed the nickname Seahorse Nebula for the cloud. Moreover, there appear to be elongated, dusty striations that are roughly perpendicularly projected with respect to the main Seahorse filament. The striations are also seen in absorption in the WISE mid-IR images (Fig. 2; cf. Fig. 3 in Miettinen 2018). Such features could be an indication that the filament is still accreting gas from its surrounding medium, but this hypothesis needs to be tested via spectral line imaging observations.

Only one of the clumps in the Seahorse IRDC appears to be associated with high-mass star formation, namely the one associated with the IR source IRAS 13039-6108. Sánchez-Monge et al. (2013) found that IRAS 13039-6108 shows radio continuum emission characteristic of an optically thin H II region. The IRAS 13039-6108 region is also surrounded by extended and diffuse IR emission, most notably seen in the WISE 12 μm image (Fig. 2), which suggests that the H II region is surrounded by a photodissociation region (PDR).

The present follow-up study of the chemistry in the Seahorse IRDC is based on observations done with the Atacama Pathfinder EXperiment (APEX2; Güsten et al. 2006). The observations and data reduction are described in Sect. 2. The analysis and the corresponding results are presented in Sect. 3. In Sect. 4, we discuss the results and in Sect. 5 we summarise our results and present our main conclusions. Throughout this paper, we adopt a distance of d = 2.54 kpc to the Seahorse IRDC (Miettinen 2012, 2018).

thumbnail Fig. 1

Herschel/SPIRE (Spectral and Photometric Imaging REceiver) 350 μm image towards the IRDC G304.74+01.32 (the Seahorse IRDC) and its surroundings. The image is displayed using a non-linear (logarithmic) stretch to improve the visibility of the faint, extended dust emission. There appear to be dusty striations on the western side of the Seahorse filament. A scale bar of 3 pc projected length is shown in the bottom left corner.
Open with DEXTER

2 Observations and data reduction

Using the Large APEX BOlometer CAmera (LABOCA; Siringo et al. 2009), the 870 μm peak positions of the clumps in the Seahorse IRDC were observed with the 12 m APEX telescope at a frequnecy of ν = 173.68831 GHz, that is, at the frequency of the SiO(J = 4−3) rotational transition. The target positions are shown in Fig. 2, and the source list is given in Table 1.

The observations were carried out on 2 September and 15–16 December 2019, and the amount of precipitable water vapour (PWV) during the observations was measured to be between 0.5 and 1.2 mm, which corresponds to a zenith atmospheric transmission range of about 96–91% at the observed frequency3.

As a front end, we used Band 5 of the Swedish ESO PI receiver for APEX (SEPIA; Belitsky et al. 2018). The SEPIA Band 5, or SEPIA180, is a sideband separating (2SB) dual polarisation receiver, which operates in the frequency range 159–211 GHz. The backend was a Fast Fourier Transform fourth-Generation spectrometer (FFTS4G) that consists of two sidebands. Each sideband has two spectral windows of 4 GHz bandwidth, which provides both orthogonal polarisations and leads to a total bandwidth of 8 GHz. The observed frequency ν = 173.68831 GHz was tuned in theupper sideband (USB). This setting allowed us to probe the frequency ranges 159.69–163.69 GHz in the lower sideband (LSB) and 171.69–175.69 GHz in the USB. The 65 536 channels of the spectrometer yielded a spectral resolution of 61 kHz, which corresponds to 105.4 m s−1 at the observed frequency of 173.68831 GHz (and ranges from 104.2 to 114.6 m s−1 over the observed frequency range). The beam size (half-power beam width; HPBW) at the observed frequency of 173.68831 GHz is 35.′′ 9 (35.′′ 5–39.′′1 over the observed frequency range), which corresponds to a physical resolution of 0.44 pc, which in turn is comparable to the diameter of the 870 μm clumps in the filament (~0.4 pc on average; Miettinen 2018).

The observations were performed in the wobbler-switching mode with a 50′′ azimuthal throw between two positions on sky (symmetric offsets), and a chopping rate of R = 0.5 Hz. The total on-source integration time (excluding overheads) was about 6.4 min for each target source. The telescope focus and pointing were optimised and checked at regular intervals on the red supergiant star VY Canis Majoris, thecarbon stars IRC +10216 (CW Leonis), 07454-7112 (RAFGL 4078), IRAS 15194-5115, and X Trianguli Australis, and the variable star of Mira Cet type R Carinae. The pointing was found to be accurate to ~ 3′′. The systemtemperatures during the observations were within the range of Tsys ≃ 77−88 K. Calibration was made by means of the chopper-wheel technique, and the output intensity scale given by the system is the antenna temperature corrected for the atmospheric attenuation (). The observed intensities were converted to the main-beam brightness temperature scale by , where ηMB is the main-beam efficiency. The value of the latter efficiency was derived as follows. The aperture efficiency at the central frequency of SEPIA180 (ηa = 0.67 when estimated from observationstowards Uranus) was first scaled to that at the observed frequency using the Ruze formula (Ruze 1952), and then converted to the value of ηMB using the approximation formula ηMB ≃ 1.2182 × ηa. This yielded the values ηa = 0.67 and ηMB = 0.824. The absolute calibration uncertainty was estimated to be about 10% (e.g. Dumke & Mac-Auliffe 2010).

The spectra were reduced using the CLASS90 of the GILDAS software package5 (version mar19b). The individual spectra or sub-scans of 9.6 s in duration (on-source time) were averaged with weights proportional to the integration time, divided by the square of the system temperature (). The resulting spectra were smoothed using the Hann window function to a velocity resolution of 210.8 m s−1 (value at 173.68831 GHz) to improve the signal-to-noise ratio (S/N). Linear (first-order) baselines were determined from the velocity ranges free of spectral line features, and then subtracted from the spectra. The resulting 1σ rms noise levels at the 0.21 km s−1 velocity resolution were about 9–17 mK on a TMB scale. The observational parameters are provided in Table 2.

thumbnail Fig. 2

WISE 12 μm image towards the IRDC G304.74+01.32 (the Seahorse IRDC). The image is displayed using a logarithmic stretch to improve the colour contrast. The overlaid contours represent the LABOCA 870 μm emission (Miettinen & Harju 2010; Miettinen 2018); the contours start at 3σ, and increase in steps of 3σ, where 3σ = 120 mJy beam−1. The target clumps are labelled so that the numbers refer to the SMM IDs (e.g. 1 refers to SMM 1), while the sources I13037, I13039, and I13042 are the three IRAS sources in the filament. The sources BLOB 1 and BLOB 2 are labelled as B1 and B2. The red plus signs indicate the target positions of the present spectral line observations (i.e. LABOCA 870 μm emissionpeaks of the clumps). A scale bar of 1 pc projected length, and the beam size at the tuned frequency of the present spectral line observations (35.′′ 9 HPBW) are shown in the bottom left corner.
Open with DEXTER
Table 1

Source sample.

Table 2

Observational parameters.

3 Analysis and results

3.1 Line identification and spectral line parameters

To identify the spectral lines in the observed frequency bands, we used the Weeds interface, which is an extension ofCLASS90 (Maret et al. 2011), to access the Cologne Database for Molecular Spectroscopy (CDMS6 ; Müller et al. 2005). The identified lines and selected spectroscopic parameters are listed in Table 3. As for the detection threshold, we used three times the rms noise level (3σ) in the spectrum. We note that all the lines were detected in the USB, and no lines were visible in the LSB, although there are lines such as C2 S(NJ = 65−56) at 161.16 GHz and C2 S(NJ = 01−12) at 166.75 GHz that one could have expected to be detectable owing to the fairly low upper-state energies of the transitions (19.2 and 9.4 K, respectively).

The spectra are shown in Figs. A.1–A.6. The SO and SiO lines were fitted by single Gaussian functions using CLASS90. All the other detected species, that is H13CN, H13CO+, HN13C, and C2H exhibit hyperfine splitting of their rotational transitions. Hence, with the exception of HN13C, the corresponding lines were fit using the hyperfine structure method of CLASS90.

The J = 2−1 transition of H13CN is split into six hyperfine components owing to the nuclear quadrupole interaction by the N nucleus (e.g. Fuchs et al. 2004). The hyperfine component frequencies were taken from the Jet Propulsion Laboratory (JPL) spectroscopic database7 (Pickett et al. 1998). The central frequency was taken to be 172.677959 GHz, which corresponds to that of the strongest hyperfine component F = 3−2 with a relative intensity of Ri = 7∕15. The H13CO+(J = 2−1) line has eight hyperfine components owing to the presence of H and 13C nuclei (Schmid-Burgk et al. 2004). To fit this hyperfine structure, we used the component frequencies from the CDMS (the strongest component, F1, F = 5∕2, 3−3∕2, 2, with Ri = 7∕20 has a frequency of 173.5067081 GHz). Although the rotational lines of HN13C are known to exhibit hyperfine splitting (e.g. Frerking et al. 1979; Turner 2001; Hirota et al. 2003; van der Tak et al. 2009; Padovani et al. 2011), to our knowledge the frequencies and relative intensities of the hyperfine components of the HN13C(J = 2−1) transition have not been published or are not available at the CDMS or the JPL database. Hence, the HN13C(J = 2−1) lines were fitted using single Gaussian profiles, which follows an approach used in previous studies and is justified by the heavy blending of the hyperfine components (e.g. Hily-Blant et al. 2010; Jin et al. 2015; Saral et al. 2018). In the C2 H molecule, the coupling of angular momentum and spin of the H nucleus leads to the hyperfine splitting of its rotational levels (e.g. Reitblat 1980; Padovani et al. 2009). We fit the detected hyperfine structure of the N = 2−1 line of C2 H using the frequencies and relative intensities of the seven components from Padovani et al. (2009; Table 5 therein). The reference hyperfine component was taken to be the strongest component J, F = 5∕2, 3−3∕2, 2 with Ri = 7∕20.

The derived line parameters are listed in Cols. 3–6 in Table A.1. Besides the formal 1σ fitting errors output by CLASS90, the errors in the peak intensity (TMB) and the integrated intensity of the line (∫ TMBdv) also include the 10% calibration uncertainty. These two sources of uncertainty were added in quadrature.

3.2 Line optical thicknesses and excitation temperatures

For spectral lines split into hyperfine components the optical thickness, τ, can be derived via the relative strengths of the different components. However, the hyperfine structure in the detected H13CO+(J = 2−1) (and HN13C(J = 2−1)) lines could not be resolved to reliably derive the corresponding optical thicknesses. In the case of H13CN(J = 2−1), the six hyperfine components were resolved into three separate groups, and hence the optical thickness could be derived via the CLASS90 hyperfine structure fitting method. However, the derived values have relatively large uncertainties owing to the partial blending of the hyperfine lines. In the case of C2 H(N = 2−1), we could clearly resolve seven hyperfine components, and hence determine the optical thickness via the aforementioned method of CLASS90. The excitation temperatures of the H13CN(J = 2−1) and C2 H(N = 2−1) transitions were then calculated by using the formula: (1)

where h is the Planck constant and the function F(Tbg) is defined as . We assumed that the background temperature, Tbg, is equal to that of the cosmic microwave background (CMB) radiation, that is Tbg = TCMB = 2.725 K (e.g. Fixsen 2009).

For all the other lines, we assumed that the rotational transition is thermalised at the dust temperature of the clump, that is we assumed that Tex = Tdust (see Col. 4 in Table 1). The corresponding line optical thicknesses were calculated by solving Eq. (1) for τ. The values of τ and Tex are listed in Cols. 7 and 8 in Table A.1.

Table 3

Detected spectral lines and selected spectroscopic parameters.

3.3 Molecular column densities and fractional abundances

To calculate the beam-averaged column densities of the molecules, N, we made an assumption of local thermodynamic equilibrium (LTE). Under this assumption, the molecular column density is given by (e.g. Turner 1991, appendix therein): (2)

where ϵ0 is the vacuum permittivity, S is the line strength, Zrot is the partition function, gK is the K-level degeneracy, gI is the reduced nuclear spin degeneracy, and the function F(Tex) is defined in a similar way as F(Tbg) in Sect. 3.2. For a Gaussian line profile, the last integral term in Eq. (2) can be expressed as: (3)

where τ0 is the peak optical thickness of the line. The total optical thickness of the lines split into hyperfine components was calculated by scaling the peak value by the relative strength of the strongest component that was always used as the reference line.

The column densities of SO, H13CN, H13CO+, SiO, and C2H were calculated from the line optical thickness as outlined above, but HN13C was treated in a different way. As mentioned in Sect. 3.1, the hyperfine component data for the HN13C(J = 2−1) transition are not published. For this reason, we assumed that the line is optically thin (τ ≪ 1), which is supported by the values of the derived peak optical thicknesses (Table A.1, Col. 7), and calculated the column density from the integrated line intensity (this approach for HN13C was also used by e.g. Hily-Blant et al. 2010). When τ ≪ 1, Eq. (1) can be solved for the velocity-integrated optical thickness as: (4)

The values of the product μ2S appearing in Eq. (2) were taken from the Splatalogue database8. All the detectedmolecular species are linear molecules, which allowed us to approximate the partition function as (e.g. Mangum & Shirley 2015; Eq. (52) therein): (5)

Equation (5) is valid at high temperatures defined as kB Tex ≫ hB (see also Turner 1991). As shown in the last three columns in Table 3, Eq. (5) yields very similar partition function values for SiO and HN13C as those given in the CDMS and JPL databases. We note that at a reference Tex of 9.375 K, the kB Tex∕hB ratios for the present species range from 4.5 to 57.8 and, hence, the aforementioned high-temperature threshold where Eq. (5) is valid is expected to be fulfilled. However, for SO and H13CN, Eq. (5) yields lower values than those in the CDMS and JPL databases (by factors of 1.7 and 3 at 9.375 K, respectively). On the other hand, for H13CO+ and C2H, Eq. (5) gives Zrot (9.375 K) values that are 12 and 2.8 times larger than in the aforementioned two databases. In the case of H13CN and C2H, the discrepancy could arise from the fact that the partition function values reported in the CDMS and JPL catalogues take the vibrational states (and potentially electronic corrections) into account (cf. Favre et al. 2014). Hence, a comparison between the rotational partition functions calculated using Eq. (5) and the partition function values reported in the CDMS and JPL catalogues is not always straightforward, and care should be taken when comparing the molecular column densities and abundances between different studies where Zrot might have been determined in different ways. Because linear molecules do not have any K-level degeneracy and the observed species do not have identical nuclei that could be interchanged while preserving the symmetry, both gK and gI are equal to unity (e.g. Turner 1991).

The fractional abundances of the molecules were calculated by dividing the column density by the H2 column density, thatis x = N∕N(H2). The N(H2) values were derived from our LABOCA dust continuum data re-reduced by Miettinen (2018). For a proper comparison with the present line observations, the LABOCA map was smoothed to the angular resolution of the line observations, where the beam size was about 1.8 times larger than in our LABOCA observations. Under the assumption of optically thin dust emission, which is valid at the wavelength probed by our LABOCA observations, N(H2) can be calculated from the dust peak surface brightness, , as (e.g. Kauffmann et al. 2008): (6)

where Bν(Tdust) is the Planck function, is the mean molecular weight per H2 molecule (taken to be 2.82, which corresponds to a hydrogen mass percentage of 71% (i.e. X = 0.71; Kauffmann et al. 2008, Appendix A.1 therein)), mH is the mass of a hydrogen atom, κν is the dust opacity, which was assumed to be 1.38 cm2 g−1 at 870 μm (based on the Ossenkopf & Henning (1994) dust model of graphite-silicate dust grains that have coagulated and accreted thin ice mantles over a period of 105 yr at a gas density of nH = 105 cm−3), and Rdg is the dust-to-gas mass ratio that we assumed to be 1/141, that is 1.41 times lower than the canonical dust-to-hydrogen mass ratio of 1/100, where the factor 1.41 (≃ 1∕X) is the ratio of total mass (H+He+metals) to hydrogen mass.

The beam-averaged column densities and abundances with respect to H2 calculated in this section are listed in the last two columns in Table A.1. Statistics of the derived fractional abundances are given in Table 4. To calculate the statistical parameters presented in Table 4, we applied survival analysis to take the left-censored data or upper limits into account. To estimate the survival curve, we used a non-parametric method called the Kaplan-Meier (K-M) estimator (Kaplan & Meier 1958). The method was implemented in R (version 3.6.2; R Core Team 2019) using the package Nondetects And Data Analysis for environmental data (NADA; Helsel 2005; Lee 2017). The distributions of the derived fractional abundances are visualised using box plots in Fig. 3. Finally, we also calculated the HN13C/H13CN and HN13C/H13CO+ abundance ratios from the corresponding column densities, and these values are listed in Table 5.

Table 4

Statistics of the molecular abundances.

thumbnail Fig. 3

Box plot (or box-and-whisker plot) showing the distributions of the derived molecular fractional abundances. The boxes are arranged in ascending order according to the median value. The height of the box represents the interquartile range IQR ≡ [Q1, Q3] (i.e. from the 25th to 75th percentile). A horizontal line that goes through the box shows the median abundance (Q2 or the 50th percentile). The whiskers are defined to extend from Q1 − 1.5 ×IQR to Q3 +1.5 ×IQR unless the minimum and maximum values of the distribution are encountered before that (in which case the whiskers extend to the minimum and maximum values). We note that only the uncensored data are used to plot the distributions (i.e. the upper limits are ignored), and SiO was detected in only one of the target sources (SMM 3).
Open with DEXTER

3.4 Correlation analysis

To search for correlations between the fractional molecular abundances and abundance ratios derived in the present study and the dust temperatures and H2 number densities listed in Table 1, we first created a correlation matrix where each cell shows the linear or Pearson correlation coefficient, r, between two variables. We note that in this analysis we used the logarithms of the abundances and densities. The relationship between two variables is generally considered strong when (e.g. Moore et al. 2018). Hence, as a threshold for potential correlation, we adopted a value of . Such correlations were found between the H13CO+ abundance and those of HN13C and C2H, and between the abundances of HN13C and C2H. The corresponding scatter plots are shown in Fig. 4.

In Fig. 4, the IR dark, IR bright, and the H II sources are shown in different colours (blue, red, and green, respectively) to better illustrate how the corresponding data points populate the plotted parameter spaces. Linear regression fits were made to the aforementioned relationships, and the derived formulas together with the r values are provided in Table 6. The y error bars were taken into account in the linear fits, but the upper limits were omitted.

4 Discussion

4.1 Detection rates and spectral line profiles

The observed spectral lines of SO, H13CO+, and HN13C were detected in all the target sources. The next highest detection rates were for C2 H (92.9%) and H13CN (85.7%). The SiO line was detected in only one of the target sources, SMM 3, which means that the corresponding detection rate was only 7.1%. The clump SMM 3 was also the only source in the Searhorse IRDC that was found to show a hint of SiO(5−4) emission (~2σ) by Miettinen (2012).

Interestingly, the SO, H13CN, H13CO+, HN13C, and C2H lines are seen in absorption towards SMM 5, SMM 6, and SMM 7. However, the H13CN line towards SMM 6 andthe H13CN and C2H lines towards SMM 7 were not found to fulfil our detection threshold of ≥ 3σ, but there is clearly a hint of line absorption, which is further supported by the positive detection of SO, H13CO+, and HN13C absorption.

A zoom-in version of Fig. 2 towards SMM 5, 6, and 7 is shown in Fig. 5. As indicated in the figure, SMM 6 is associated with an embedded young stellar object (YSO) seen in the mid-IR. Hence, the detected absorption lines might be caused by the outer layers of the clump absorbing the millimetre-wavelength continuum photons emitted by the warm or hot (~ 100 K; Rathborne et al. 2011) dust close to the central YSO (e.g. Doty & Neufeld 1997). There is also an IR source detected with WISE only at 22 μm towards SMM 7, but the source, J130704.50-612620.6, is likely a chance projection of a star-forming galaxy or an active galactic nucleus (Miettinen 2018). This extragalactic object seen towards SMM 7 could act as the continuum source against which the absorption lines are generated by the clump medium. The clump SMM 5 appears to not have any IR point sources within the ~ 36′′ beam of the present line observations, and hence the detection of absorption lines towards it is somewhat surprising. As can be seen in Fig. 2, SMM 5 is associated with an elongated IR dark feature that could be associated with the striation on its western side. In principle, the system’s geometry could be such that the dense gas associated with the striation acts as a foreground absorbing layer for the sight-line towards SMM 5, but this is highly speculative at present. Spectral line imaging of the Seahorse and its surroundings are needed to quantitatively study these issues.

We note that the C17O(2−1) lines were detected in emission towards SMM 5, SMM 6, and SMM 7 by Miettinen (2012; Fig. 2 therein), but the beam size of those observations was 1.3 times smaller than that of the present observations, and the target positions towards the aforementioned clumps were offset from the present ones by 5.′′ 7–25.′′6 because they were based on an earlier version of our LABOCA map and different method of clump extraction (Miettinen & Harju 2010;Miettinen 2018).

Aside from being visible in absorption, the H13CO+ and HN13C lines towards SMM 6 exhibit a double-peaked profile with a stronger blueshifted peak compared to the redshifted peak. Such blue asymmetric profiles were also detected by Miettinen (2012) in the 13CO(2−1) lines towards clumps in the southern half of the Seahorse IRDC, most notably towards (SMM 2; Fig. 3 therein), and they are indicative of large-scale infall gas motions. This supports the hypothesis that the Seahorse IRDC is still accreting mass from its surrounding gas reservoirs (Sect. 1).

Finally, the spectral line profile of H13CO+ towards SMM 3 shows an absorption-like feature next to the actual spectral line, while the H13CO+ and HN13C absorption lines towards SMM 5, H13CN absorption line towards SMM 6 (below the detection threshold), and the C2 H absorption lines towards SMM 5 and SMM 6 exhibit emission-like features next to the actual lines. These are indications of the presence of H13CN, H13CO+, HN13C, and C2 H gas in the observations’ off-position (50′′ azimuthal offset from the target position). If there is molecular gas emission in the observations’ off-position, then, owing to the wobbler switching technique, when the off-spectrum is subtracted from the on-source spectrum, one can see an absorption-like feature in the final spectrum. Similarly, an absoprtion line in the off-position can appear as an emission-like feature in the final spectrum.

thumbnail Fig. 4

Scatter plots between the variables that were found to be linearly correlated with a Pearson correlation coefficient of r > 0.7. The arrows pointing down indicate upper limits. The dashed lines show the linear regression fits to the uncensored data points (i.e. the upper limits were not taken into account).
Open with DEXTER
Table 5

HN13C/H13CN and HN13C/H13CO+ abundance ratios.

Table 6

Equations of the linear regressions plotted in Fig. 4 and the corresponding correlation coefficients.

thumbnail Fig. 5

Zoom-in of Fig. 2 towards SMM 5, 6, and 7 for which some of the spectral lines are seen in absorption. Only the 3σ and 6σ LABOCA contours are plotted for clarity. Besides the clumps SMM 5, 6, and 7, also the striation seen as a dark absorption feature is labelled (cf. Fig. 1, where the striation is seen in dust emission). The red plus signs indicate the target positions of the present spectral line observations, while the yellow diamond symbols show the positions of the WISE sources seen towards SMM 6 and SMM 7. The latter one is likely to be an extragalactic object, thatis a star-forming galaxy or an active galactic nucleus, and hence SMM 7 is considered an IR dark clump (Miettinen 2018). A scale bar of 0.5 pc projected length, and the beam size at the tuning frequency of the present spectral lineobservations are shown in the bottom left corner.
Open with DEXTER

4.2 Relationships between the molecular abundances

We found three potential cases of correlation between fractional abundances of two molecular species (Fig. 4). Below, we discuss each of these in the order of decreasing significance (as quantified by the Pearson correlation coefficient).

The strongest correlation was found between the abundances of HN13C and C2 H (r = 0.93). According to the gas-phase chemistry models, the most abundant isotopologue of HN13C, HNC, is primarily formed through the dissociative recombination reaction HCNH+ + e−→HNC + H (e.g. Herbst 1978). Other dissociative recombination reactions that can form HNC are H2 CN+ + e− and H2 NC+ + e− (Pearson & Schaefer 1974; Allen et al. 1980). The neutral-neutral reaction NH2 + C →HNC + H can also act as the source of HNC (e.g. Herbst et al. 2000). In the PDRs, C2 H can form as a result of the photodissociation of C2 H2 (C2 H2 + hν →C2H + H; e.g. Fuente et al. 1993). The C2 H molecules can also form via the dissociative recombination reaction (e.g. Turner et al. 2000). Hence, if C2H is predominantly formed via the aforementioned dissociative recombination reaction, the correlation we found between HN13C and C2 H could reflect the gas-phase electron abundance (i.e. ionisation degree). If C2 H is mainly formed via the photodissociation of C2 H2, the positive relationship with HN13C could be linked to photoionisation that yields free electrons from which HN13C can form. On the other hand, when ultraviolet (UV) photons destroy CO via photodissociation (e.g. Visser et al. 2009), the resulting C atoms can react with CH2 to form C2H (e.g. Turner et al. 2000). The aforementioned correlation could therefore also reflect the gas-phase abundance of C atoms (and hence the importance of UV photodissociation in the formation of both HNC (and HN13C) and C2H).

With a correlation coefficient of r = 0.88, the correlation between the H13CO+ and HN13C abundances was found to be the second strongest of those plotted in Fig. 4. The HCO+ molecules are predominantly formed in the reaction (e.g. Herbst & Klemperer 1973), where the required CO can form in the reaction C + OH →CO + H (e.g. Millar 2015). Ionised carbon can also react with water to form HCO+ (C + + H2O →HCO+ + H; e.g. Rawlings et al. 2004). The 13C isotopologue of HCO+ we observed, H13CO+, forms in a similar was as HCO+ but from 13CO. The isotope exchange reaction HCO+ + 13CO →H13CO+ + CO provides another route for the formation of H13CO+ (e.g. Langer et al. 1984; Mladenović & Roueff 2017). The correlation we found between H13CO+ and HN13C could therefore be a manifestation of their dependence on carbon. However, the chemical network is complicated by the fact that HCO+ is mainly destroyed via the dissociative recombination HCO+ + e−→CO + H (e.g. Liszt 2017). As described above, dissociative recombination reactions can also lead to the formation of HNC, and because the reaction HCO+ + e− is the main source of CO from which HCO+ can re-form, our finding is also likely to reflect the degree of ionisation in the gas.

The positive correlation we found between the H13CO+ and C2H abundances (r = 0.74) is to be expected from the aforementioned two relationships. However, the slope of the linear least squares fit shown in the middle panel in Fig. 4 deviates from a flat trend by only 1.5σ. On the other hand, the upper limit to the C2H abundance at the lowest H13CO+ abundance, which was not used in the fit, supports the presence of a positive correlation.

4.3 Comparison of the molecular abundances in the Seahorse IRDC to those in other IRDCs

4.3.1 SO

Observations of SO spectral lines towards IRDCs are quite rare. Sakai et al. (2010) used the Nobeyama Radio Observatory (NRO) 45 m telescope to observe the SO(NJ = 22 − 11) line at 86 GHz towards 20 clumps in different IRDCs. The detection rate was 70%. The authors reported the abundances with respect to H13CO+ (their Table 5), so we decided to compare the molecular column densities instead of the fractional abundances with respect to H2. The N(SO) distribution derived in the present work is compared with that from Sakai et al. (2010) in the top left panel in Fig. 6. All our values are lower than those derived by Sakai et al. (2010), and our mean (median) SO column density is lower by a factor of 53 (42).

Sanhueza et al. (2012) used the Mopra 22 m telescope to observe the SO(22−11) line towards a sample of 92 clumps in IRDCs. However, the line was detected in only <8 clumps, and hence the SO properties were not analysed further in their statistical study. To our knowledge, the present detections of SO(NJ = 44−33) line emission at 172 GHz towards an IRDC is the first of its kind.

Through observations of SO(NJ = 65−54) towards 59 high-mass star-forming objects in total, Gerner et al. (2014) derived median SO abundances of <1.3 × 10−10, 9 × 10−10, 9 × 10−10, and 3.8 × 10−9 for their IRDCs, high-mass protostellar objects (HMPOs), hot molecular cores (HMCs), and ultracompact (UC) H II regions, respectively (assuming excitation temperatures of 15, 50, 100, and 100 K, respectively; see their Tables 3–6). The authors assumed that the gas-to-dust mass ratio is 100, 870 μm dust opacity is 1.42 cm2 g−1, and that the mean molecular weight is 2.8. Hence, we scaled down their reported molecular abundances by a factor of 0.6942 to take these different assumptions into account. The aforementioned value for the IRDCs is comparable (although it is an upper limit) to the present median SO abundance of 1.5 × 10−10 for IR dark clumps, but we found a factor of 8.2 lower median SO abundance for the IR bright clumps than in the HMPOs and HMCs in the Gerner et al. (2014) sample. Finally, the one H II region in the Seahorse IRDC exhibits a x(SO) value that is over 40 times lower than the UC H II regions’ median x(SO) in the Gerner et al. (2014) sample.

thumbnail Fig. 6

Comparison of the molecular abundance distributions (column density for SO) in the Seahorse IRDC with those in other IRDC sources presented in the literature. The box plots are sorted by increasing median abundance. The data points above some of the boxes represent outlier values. Upper limits are neglected from the distributions.
Open with DEXTER

4.3.2 H13CN

The H13CN(1−0) transition was part of the Mopra 90 GHz molecular line survey of 15 IRDCs (37 clumps in total) by Vasyunina et al. (2011). The line was detected in only five sources (13.5% detection rate; their Table 3). The corresponding abundances are compared with the present values in the top middle panel in Fig. 6. We note that Vasyunina et al. (2011) adopted a gas-to-dust mass ratio of 100, and assumed that the dust opacity is 1 cm2 g−1 at their observed dust emission wavelength of 1.2 mm. Moreover, a hydrogen molecule mass was used in the denominator of their N(H2) formula, which effectively means that the mean molecular weight was 2 (see Eq. (2) in Vasyunina et al. 2009). Hence, we scaled down their molecular abundances by a factor of 0.7735 for a more meaningful comparison with our values (the dust emissivity index, β, used to scale the dust opacity as κν ∝ νβ was assumed to be β ≃ 1.8; see Miettinen 2018, and references therein). The H13CN abundances in the Seahorse IRDC are similar to those in the Vasyunina et al. (2011) sample. For example, our mean (median) value is only 1.1 (0.6) times their value. However, we note that Vasyunina et al. (2011) calculated the rotational partition functions byinterpolating data from the CDMS catalogue values, which complicates our comparison as discussed in Sect. 3.3. Hence, part of the discrepancy can still arise from the different methods of calculation rather than being physical (or the true difference could be larger than quoted above).

As in the case of SO, Sanhueza et al. (2012) observed the H13CN(1 − 0) line but detected it in <8 sources, and hence no quantitative comparison could be made with their study. Our detection rate of H13CN towards the Seahorse IRDC, 85.7%, seems quite high compared to the studies by Vasyunina et al. (2011) and Sanhueza et al. (2012), although the mean abundance of the molecule in the Seahorse appears to be very similar to that derived for the Vasyunina et al. (2011) sources.

4.3.3 H13CO+

The H13CO+ species has commonly been included in the previous studies of chemistry in the IRDCs (e.g. Sakai et al. 2010; Vasyunina et al. 2011; Sanhueza et al. 2012; Miettinen 2014; Gerner et al. 2014). The detection rate of this species has also been high in some of the previous surveys, most notably 100% in the surveys carried out by Sakai et al. (2010) and Gerner et al. (2014), which is also the case in the present study.

Sakai et al. (2010) reported H13CO+ column densities in the range of (1.3−13.5) × 1012 cm−3 with a mean (median) of (6.2 ± 0.8) × 1012 cm−3 (6.0 × 1012 cm−3). The H13CO+ column densities we derived, (1.1−11) × 1012 cm−3 with a mean (median) of (6.3 ± 0.8) × 1012 cm−3 (6.2 × 1012 cm−3) are remarkably similar to those from Sakai et al. (2010).

In the top right panel in Fig. 6, we visualise the comparison of the H13CO+ abundances with those from previous studies. Again, the values from Vasyunina et al. (2011) were multiplied by 0.7735 because of the different assumptions used in the analysis (Sect. 4.3.2). Sanhueza et al. (2012) made the same assumptions about the dust properties (opacity and amount with respect to gas) as Vasyunina et al. (2011). However, no information about the mean molecular weight used in the calculation of N(H2) was given. Nevertheless, from the reported 1.2 mm surface brightnesses and the corresponding N(H2) values we derived a mean molecular weight similar to that assumed in the present study. Hence, the fractional abundances reported by Sanhueza et al. (2012; Table 10 therein) were scaled down by a factor of 0.549 for a proper comparison with our results. Moreover, we note that Sanhueza et al. (2012) reported the HCO+ abundancies (rather than H13CO+), which were divided by 50 to obtain the H13CO+ abundances (the authors assumed an [HCO+]/[H13CO+] abundance ratio of 50 for all sources). Miettinen (2014) assumed a same dust opacity value as we did, but the assumed dust-to-gas mass ratio and mean molecular weight were different (1/100 and 2.8, respectively). Hence, the abundances from Miettinen (2014) were multipled by 0.714 to bring them to level where a direct comparison is meaningful. The H13CO+ abundances in the Seahorse IRDC are in the same ballpark with those in other IRDCs. For example, the mean (median) abundances for the Vasyunina et al. (2011) and Sanhueza et al. (2012) samples are (2.7 ± 0.3) × 10−10 (2.3 × 10−10) and (4.3 ± 0.5) × 10−10 (2.8 × 10−10), while our corresponding values are (6.2 ± 1.0) × 10−10 (6.1 × 10−10).

Gerner et al. (2014) used H13CO+(J = 1 − 0) observations, and derived the median abundances of 3.1 × 10−11, 3.1 × 10−10, 5.7 × 10−10, and 6.9 × 10−10 for IRDCs, HMPOs, HMCs, and UC H II regions, respectively. Again, their abundances were scaled down for a better comparison with the present results (Sect. 4.3.1). We derived about 27 times higher median H13CO+ abundance for our IR dark clumps than in the aforementioned IRDCs, but our median H13CO+ abundance for IR bright clumps (5.7 × 10−10) is very similar to those in the HMPOs and HMCs observed by Gerner et al. (2014). Also, the H13CO+ abundance in our H II region, (6.8 ± 0.8) × 10−10, is in excellent agreement with their median abundance in UC H II regions.

Saral et al. (2018) studied 30 IRDC clumps that were drawn from the 870 μm APEX Telescope Large Area Survey (ATLASGAL; Schuller et al. 2009) and that appear dark at 8 μm or 24 μm wavelengths. The mean (median) H13CO+ column density reported by the authors, (1.8 ± 0.1) × 1012 cm−3 (1.4 × 1012 cm−3), is 3.5 (4.4) times lower than the present value, which could simply reflect the aforementioned selection effect, namely that the Saral et al. (2018) sample is solely composed of mid-IR dark sources. On the other hand, our IR dark clumps have a mean H13CO+ column density of (6.7 ± 1.7) × 1012 cm−3, which is very similar to our full sample mean of (6.3 ± 0.8) × 1012 cm−3.

4.3.4 SiO

We detected SiO in only one of the target clumps, SMM 3. This source was also the only one of the five target sources in the Seahorse where SiO(J = 5−4) was tentatively detected (~2σ) by Miettinen (2012). In the bottom left panel in Fig. 6, we plot the SiO abundance distributions from previous studies and compare them with the SiO abundance derived for SMM 3. The abundances from Vasyunina et al. (2011), Sanhueza et al. (2012), and Miettinen (2012) were scaled as described above. Sanhueza et al. (2012) employed the same assumptions in their analysis as Sanhueza et al. (2012), and hence their SiO abundances were scaled down by a factor of 0.549. Clearly the SiO abundance in SMM 3 appears to be very low compared to the typical values in clumps in other IRDCs. For example, the mean x(SiO) values in the aforementioned reference studies are 4–9.7 times higher than a value of x(SiO) = 7.4 × 10−11 in SMM 3.

Gerner et al. (2014) observed the J = 2−1 transition of SiO, and derived the median abundances of 5.8 × 10−11, 2.8 × 10−10, 9.0 × 10−10, and 2.6 × 10−10 for IRDCs, HMPOs, HMCs, and UC H II regions, respectively (where the quoted values were scaled to the present assumptions). The SiO abundance we derived for the IR bright clump SMM 3 is as low as observed by Gerner et al. (2014) in IRDCs, but about 4–12 times lower than the median abundance in their HMPOs and HMCs.

Li et al. (2019) carried out an SiO(J = 5−4) survey of a sample of 44 massive clumps in IRDCs, where the detection rate was 57%. The mean (median) SiO abundance reported by the authors is 3.7 × 10−11 (2.6 × 10−11), which was scaled down by a factor of 0.64 to take into account the different assumptions in the analysis (Li et al. (2019) used the values Rdg = 1∕100, κ870,μm = 1.54 cm2 g−1, and ). This is within a factor of 2 (2.8) of the SiO abundance in SMM 3.

Both Sakai et al. (2010) and Saral et al. (2018) determined SiO column densities in IRDC clumps. In the former study, the mean (median) SiO column density was (1.5 ± 0.2) × 1013 cm−3 (1.3 × 1013 cm−3), while in the latter one it was (2.0 ± 0.1) × 1012 cm−3 (1.4 × 1012 cm−3). The derived value of N(SiO) towards SMM 3, (1.0 ± 0.1) × 1012 cm−3, is more similar to the mean value from Saral et al. (2018), and 15 times lower than the mean for the Sakai et al. (2010) sample.

Because we have observed a fairly high-J transition of SiO, J = 4−3, whose upper-state energy is 20.84 K the low detection rate could be explained by the transition not being excited at the physical conditions of the Seahorse IRDC. This hypothesis could be tested through observations of low-J transitions of SiO (preferably J = 1−0; Eu ∕kB = 2.08 K). In particular, mapping observations of low-J SiO line emission would be useful to explore the possibility that the Seahorse filament is formed through converging, supersonic turbulent flows. In some other filamentary IRDCs, detection of widespread (≳ 2 pc) SiO emission is suggested to be a potential manifestation of the large-scale shock associated with the cloud formation process via colliding turbulent flows (Jiménez-Serra et al. 2010; Cosentino et al. 2018). For example, Jiménez-Serra et al. (2010) mapped the filamentary IRDC G035.39-00.33 in the J = 2−1 transition of SiO (Eu ∕kB = 6.25 K), and single-pointing observations were also done in the J = 3−2 transition of SiO (Eu ∕kB = 12.50 K). The SiO abundance derived for the narrow-line (~ 0.8 km s−1) component in G035.39-00.33 (only detected in SiO(2−1)) was found to be ~10−11, and, as one potential explanation, the authors suggested that such SiO emission could have its origin in the processing of dust grains by a shock associated with the formation of the IRDC. As pointed out by the authors, the narrow-line SiO emission could also be caused by outflows from YSOs like the broader lines (~ 4−7 km s−1) they detected, but possibly caught in a decelerated or recently activated stage.

The SiO(4−3) line we detected towards SMM 3 is broad, 8.30 ± 0.50 km s−1 in FWHM, and is very likely to arise from a YSO outflow-driven shock. However, the SiO abundance we derived for SMM 3, (7.4 ± 0.9) × 10−11, appears lower than those derived by Jiménez-Serra et al. (2010) towards YSOs in G035.39-00.33 (~ 10−10 to ≥ 10−8). Also, the upper limits to the SiO abundance we estimated along the Seahorse IRDC, ranging from < 5 × 10−12 to < 2.2 × 10−11, are even lower than expected from a large-scale shock if the filament was formed in the same way as G035.39-00.33 and is of comparable age. However, Jiménez-Serra et al. (2010) calculated the SiO column densities using the large velocity gradient modelling, and derived the abundances with respect to N(H2) estimated from CO isotopologues, which makes a direct comparison with their results difficult. Again, observations of Jup < 4 transitions of SiO towards the Seahorse are needed to shed light on these issues.

4.3.5 HN13C

In the bottom middle panel in Fig. 6, the derived HN13C abundances are compared with those derived by Sanhueza et al. (2012) and Miettinen (2014), where the aforementioned scaling factors were again applied. Sanhueza et al. (2012) detected HN13C in 34% of their target clumps, and to obtain their HN13C abundances we divided the reported HNC abundances by 50 (the authors assumed an [HNC]/[HN13C] abundance ratio of 50 for all their sources). The mean (median) value of x(HN13C) for the Sanhueza et al. (2012) sample is (5.3 ± 0.4) × 10−10 (4.1 × 10−10), while for the present sample it is (2.4 ± 0.3) × 10−11 (2.5 × 10−11), that is over 20 (16) times lower than found by Sanhueza et al. (2012).

Gerner et al. (2014) observed the HN13C(J = 1−0) line towards their sample, and their scaled-down median HN13C abundances for IRDCs, HMPOs, HMCs, and UC H II regions are 4.0 × 10−11, 2.6 × 10−10, 3.4 × 10−10, and 5.3 × 10−10, respectively. We derived a median HN13C abundance of a few times 10−11 in every source types in our sample (IR dark and bright clumps, and an H II region), which is comparable to that found by Gerner et al. (2014) in their IRDCs, but over an order of magnitude lower for clumps clearly associated with YSOs.

Column densities of HN13C towards IRDCs were determined by Sakai et al. (2010) and Saral et al. (2018). The mean (median) values for their samples are (7.8 ± 0.8) × 1012 cm−3 (7.0 × 1012 cm−3) and (3.0 ± 0.1) × 1012 cm−3 (1.6 × 1012 cm−3). Compared to the present value, (2.4 ± 0.3) × 1011 cm−3 (2.1 × 1011 cm−3), those values are over an order of magnitude higher.

4.3.6 C2H

As illustrated in Fig. 3, C2H was found to be the most abundant molecule in the present study, but we note that together with H13CN it was also the only case where the optical thickness and hence excitation temperature could be derived via hyperfine structure fitting. An abundance comparison with previous studies (Vasyunina et al. 2011; Sanhueza et al. 2012; Miettinen 2014) is visualised inthe bottom right panel in Fig. 6. The reference abundances were scaled down as discussed above.

The C2H abundances in the Seahorse appear to be lower than in many other IRDCs. The mean (median) abundances for the Vasyunina et al. (2011), Sanhueza et al. (2012), and Miettinen (2014) samples are, respectively, (1.1 ± 0.2) × 10−8 (8.9 × 10−9), (2.6 ± 0.3) × 10−8 (2.2 × 10−8), and (5.8 ± 1.2) × 10−9 (4.5 × 10−9). The mean (median) C2 H abundance in the Seahorse, (2.3 ± 0.3) × 10−9 (2.5 × 10−9), is 4.8 (3.6), 11.3 (8.8), and 2.5 (1.8) times lower than the aforementioned values, respectively.

The median C2H abundances in the IRDCs, HMPOs, HMCs, and UC H II regions studied by Gerner et al. (2014) were derived to be 8.3 × 10−9, 9.7 × 10−8, 2.2 × 10−7, and 2.7 × 10−7, respectively.These are all larger than those we derived for the Seahorse clumps, by a factor of 3.1 for the IR dark clumps, factors of ~ 40−92 for the clumps associated with YSOs, and by a factor of over 190 when comparing the H II region values.

The mean (median) C2H column density we derived, (1.3 ± 0.2) × 1015 cm−3 (1.1 × 1015 cm−3), is almost an order of magnitude higher than those derived by Sakai et al. (2010; (1.5 ± 0.2) × 1014 cm−3 (1.4 × 1014 cm−3)).

4.3.7 Abundance ratios

In the present study, we also derived the [HN13C]∕[H13CN] and [HN13C]∕[H13CO+] abundance ratios (see Table 5). The full sample means are 0.30 ± 0.07 and 0.04 ± 0.005, respectively. If the 13C fractionation is equally (un-)important or (in-)efficient in the molecules of the aforementioned ratios, then the ratios are expected to be equal to [HNC]∕[HCN] and [HNC]∕[HCO+] (e.g. Roberts et al. 2012).

The [HNC]∕[HCN] ratio in IRDCs is found to be about unity on average (e.g. Vasyunina et al. 2011; Liu et al. 2013; Miettinen 2014; Jin et al. 2015), which is consistent with theoretical expectations for dark interstellar clouds (e.g. Sarrasin et al. 2010; Loison et al. 2014). For example, the mean [HNC]∕[HCN] ratio and its standard error for the IRDC clump sample of Miettinen (2014) is 1.26 ± 0.46, which is 4.2 ± 1.8 times higher than the present mean [HN13C]∕[H13CN] ratio. Hence, the [HN13C]∕[H13CN] ratio in the Seahorse IRDC appears to be lower than expected for the 12C isotopologues of the observed species.

On the other hand, the [HNC]∕[HCN] ratio is subject to temperature variations. For example, the [HNC]∕[HCN] ratio in Orion was found to increase from 0.003–0.005 in the hot plateau and hot core regions (~ 150−200 K) to 0.017–0.17 in the colder(~30−60 K) quiescent parts of the cloud (Goldsmith et al. 1986; see also Schilke et al. 1992; Hacar et al. 2020). Hirota et al. (1998) found that the [HNC]∕[HCN] ratio drops rapidly when the gas kinetic temperature becomes Tkin > 24 K, that is at temperatures higher than typically found in dark clouds (~ 10 K). The authors suggested that the reason why the [HNC]∕[HCN] ratio drops is that HNC is being destroyed by neutral-neutral reactions some of which can form HCN (HNC + H →HCN + H and HNC + O →NH + CO). We note that the neutral-neutral reaction CH2 + N →HCN + H could also play a role in lowering the [HNC]∕[HCN] ratio (e.g. Herbst et al. 2000). However, Hirota et al. (1998) did not found any significant differences in [HNC]∕[HCN] between star-forming cores and starless cores, which is in reasonable agreement with our comparable mean [HN 13C]∕[H13CN] ratios derived for the IR dark and IR bright clumps (0.38 ± 0.11 and 0.27 ± 0.10, respectively, whose ratio is 1.4 ± 0.7). However, the median [HN13C]∕[H13CN] ratios derived for our IR dark and IR bright clumps show a more significant difference (being 2.8 times higher in the IR dark clumps). In the H II region clump IRAS 13039-6108, the [HN 13C]∕[H13CN] ratio was derived to be 0.05 ± 0.03, which is lower than the averages in the IR dark and IR bright sources by factors of 7.6 ± 5.1 and 5.4 ± 3.8, respectively. The potential evolutionarytrend in the [HN13C]∕[H13CN] ratio in the Seahorse is further discussed in Sect. 4.4.

Gerner et al. (2014) modelled the chemical evolution of different stages of high-mass star formation using a one-dimensional physical model of density and temperature evolution coupled with a time-dependent gas-grain chemical model. Their best-fit model followed the aforementioned trend observed in Orion, and the [HNC]∕[HCN] ratio reached a minimum of ~0.02 at the beginning of the hot core stage (see their Fig. 6). However, their observed [HNC]∕[HCN] ratio did not exhibit such a clear trend, but instead the ratio was found to be comparable for all the four types of sources they studies, namely IRDCs, HMPOs, HMCs, and UC H II regions (all values were in the range ~ 1.−3.3).

Jin et al. (2015) found that the mean [HNC]∕[HCN] ratio drops from about unity in quiescent IRDC cores (neither 4.5 μm nor 24 μm emission) to ~ 0.4 in active IRDC cores (associated with both 4.5 and 24 μm emission) to ~0.2 in HMPOs to ~0.1 in UC H II regions. This behaviour can be understood in terms of gas temperature increase with source evolution as discussed above. On the other hand, we did not find any correlation between [HN 13C]∕[H13CN] and Tdust. The dust temperatures in the Seahorse clumps might be too similar to each other for the temperature-dependent [HNC]∕[HCN] evolutionary indicator to be manifested (the dust temperatures range from 11.5 to 22.2 K; Table 1).

Colzi et al. (2018) observed the J = 1−0 transition of different isotopologues of HCN and HNC towards 27 high-mass star-forming objects including high-mass starless cores (HMSCs), HMPOs, and UC H II regions. The authors used a Galactocentric dependent ratio to convert the column densities of H13CN and HN13C to those of the main isotopologues. and found that the mean [HNC]∕[HCN] ratios for HMSCs, HMPOs, and UC H II regions are 1.3 ± 0.3, 0.4 ± 0.03, and 0.5 ± 0.1, respectively.For the first evolutionary stages, the results obtained by Colzi et al. (2018) are similar to those from Jin et al. (2015), but the mean [HNC]∕[HCN] ratio derived by Colzi et al. (2018) for their UC H II regions does not appear to drop compared to HMPOs as found by Jin et al. (2015). Finally, we note that Saral et al. (2018) derived a high mean [HNC]∕[HCN] ratio of 8.9 ± 1.0 for their sample(median was 7.1), which could be caused by their sample being, by construction, biased towards clumps in very early stages of evolution (see Sect. 4.3.3).

The average [HNC]∕[HCO+] ratios in IRDCs are found to be ~2 (e.g. 1.7 ± 0.2 for the Liu et al. 2013 sample, and 2.3 ± 0.1 for the Saral et al. 2018 sample). This is in fair agreement with the gas-phase chemical models of cold dark clouds, which suggest comparable abundances for these species (e.g. Roberts et al. 2012 and references therein). However, Miettinen (2014) derived a higher mean value of 4.3 ± 0.7 (median was 2.8). Our mean [HN13C]∕[H13CO+] ratio is ~50−108 times lower than the main isotopologue ratio in the aforementioned studies. At least partly, this could be assigned to 13C fractionation effects being different for the two molecules. Miettinen (2014) derived comparable mean abundances for HN13C and H13CO+, their ratio being 1.2, which is 30 times higher than our mean ratio. On the other hand, Saral et al. (2018) suggested that the [HNC]∕[HCO+] ratio could be subject to environmental effects (density, turbulence; Godard et al. 2010), and hence an unreliable source evolutionary indicator. Observations of the main isotopologues HNC and HCO+ towards the Seahorse are required to examine how the [HNC]∕[HCO+] ratio compares with those in other Galactic IRDCs, and to quantify the 13C fractionation differences in the species.

4.4 Potential evolutionary indicators

Because our sample is composed of different types of clumps, namely five IR dark clumps, eight IR bright clumps, and one clump associated with an H II region, we searched for potential evolutionary trends among the different physical and chemical parameters discussed in the present paper. The four possible trends that we found are shown in Fig. 7. We note that although our sample is small, and hence subject to small number statistics, our results benefit from the fact that all the sources belong to the same parent cloud, and hence lie at the same distance and are expected to share similar initial chemical and physical conditions. Thanks to a common distance between all the sources, distance-dependent beam-dilution effects are expected to play no role in the derived chemical abundances (or rather the effect is similar for all sources).

As shown in the top left panel in Fig. 7, we found a hint that the H13CN abundance increases as the clump evolves from the IR dark stage to IR bright and H II region stages. The mean (median) abundances derived for these three types of sources are (2.7 ± 2.2) × 10−10 (0.7 × 10−10), (4.0 ± 1.8) × 10−10 (1.0 × 10−10), and (4.1 ± 2.4) × 10−10 (only one source). These are not statistically significant differences and, as noted above, we are dealing with a fairly small sample, especially a sample of one for the H II regions. Nevertheless, our result is consistent with the increasing trend in the HCN abundance from an IRDC stage to the ignition of a HMC found by Gerner et al. (2014) in their observational data and through chemical modelling (see their Fig. 5). Also, Jin et al. (2015) found that the peak intensity of the sub-sample averaged H13CN spectra gets stronger from quiescent IRDC cores to UC H II regions. As discussed in Sect. 4.3.7, an increasing evolutionary trend in x(H 13CN) can be understood to arise from the neutral-neutral reactions HNC + H →HCN + H and CH2 + N →HCN + H being favoured by the increasing abundances of hydrogen and nitrogen atoms as the gas and dust temperature of the source rises.

In the top right panel in Fig. 7, a decreasing trend in the [HN 13C]∕[H13CN] ratio is seen as the clump evolves. The mean (median) ratio in the three types of clumps was found to decrease from 0.38 ± 0.11 (0.54) in the IR dark clumps to 0.27 ± 0.10 (0.19) in the IR bright clumps, and further to 0.05 ± 0.03 in the H II region stage. This behaviour is consistent with the positive evolutionary trend of x(H 13CN) shown in the top left panel in Fig. 7, and was already addressed in Sect. 4.3.7.

The decreasing evolutionary trend in the C2H abundance is shown in the bottom left panel in Fig. 7. The mean (median) x(C2H) drops from (2.7 ± 0.4) × 10−9 (2.7 × 10−9) to (2.2 ± 0.3) × 10−9 (2.4 × 10−9) and to (1.4 ± 0.2) × 10−9 for our three types of clumps. Sakai et al. (2010) concluded that the C2 H abundance (with respect to H13CO+) does not enhance significantly when star formation in the clump kicks in. Sanhueza et al. (2012) found that N(C2H) increases with clump evolution, but such trend is not visible in their x(C2H) values (Table 8 therein). On the other hand, Miettinen (2014) found that both the mean and median values of x(C2H) are higher in IRdark clumps than in IR bright clumps, which is similar to our result.

Gerner et al. (2014) observed an increasing trend in the C2 H abundance when the source evolves from an IRDC stage to a HMPO, and further to a HMC and then to an UC H II region (see their Fig. 5). However, their modelled values showed a different behaviour so that x(C2H) first increased from IRDCs to HMPOs, but then dropped for HMCs and UC H II regions. The behaviour at the beginning is opposite to our finding, but the later decreasing evolutionary trend resembles our observational results. The latter is consistent with a suggestion that the C2H abundance decreases in the hot core stage of evolution (Beuther et al. 2008). For example, C2 H molecules in hot cores react rapidly with O atoms to form CO (C2H + O →CO + CH; e.g. Watt et al. 1988). Furthermore, there is observational evidence that C2 H is destroyed in H II regions (e.g. Walsh et al. 2010); the far-UV photons can dissociate C2 H into H atoms and C2 molecules (e.g. Nagy et al. 2015). This could also explain the low value of x(C2H) observed in IRAS 13039-6108.

Finally, we found that the volume-averaged H2 number density appears to decrease as the clump evolves (Fig. 7, bottom right panel). The mean (median) density drops from (5.2 ± 0.4) × 104 cm−3 (5.1 × 104 cm−3) to (4.0 ± 0.4) × 104 cm−3 (4.9 × 104 cm−3) and further to (3.0 ± 1.6) × 104 cm−3 for the three stages of clump evolution. This is a counter-intuitive result because one would expect the density to increase as the clump contracts under the influence of gravity. Previous studies of high-mass clumps have indeed demonstrated that the average densities of star-forming clumps are higher than those that show no signs of star formation (e.g. Chambers et al. 2009; Giannetti et al. 2013). Although one can think of multiple sources of uncertainty in the H2 number density calculation, such as assuming a spherical geometry for clumps embedded in a filamentary parent cloud and using the same dust opacity for all types of clumps in calculating their mass and density (), it is still puzzling to see a decreasing trend shown in Fig. 7. One could speculate that the H2 content of the Seahorse is complicated by its potential accretion of gas from the surrounding medium and stellar feedback effects (H2 destruction by UV radiation in star-forming clumps; e.g. McKee & Krumholz 2010), but more detailed studies are needed to test such speculations.

thumbnail Fig. 7

Box plots showing the clump properties that were found to be different on average for the different types of clumps in the present sample, namely clumps that are either dark or bright in the mid-IR and clumps associated with an H II region (onlyone source, IRAS 13039-6108). Censored data are not included in the plot. Some of the trends could be manifestations of clump evolution (Sect. 4.4).
Open with DEXTER

5 Summary and conclusions

We used theAPEX telescope to carry out a molecular line survey at 170 GHz towards the clumps in the filamentary IRDC G304.74+01.32, nicknamed the Seahorse Nebula. The new spectral line data were combined with our previous LABOCA 870 μm dust continuum data to calculate the fractional abundances of the detected molecules. Our main results are summarised as follows:

  • 1.

    Altogether six different molecular species were detected, namely SO, H13CN, H13CO+, SiO, HN13C, and C2H. The species SO, H13CO+, and HN13C were detected in all the 14 target sources. The next highest detection rates were for C2H (92.9%) and H13CN (85.7%). Only one source (SMM 3) was found to be an SiO emitter at ~174 GHz (7.1% detection rate).

  • 2.

    The SO, H13CN, H13CO+, HN13C, and C2H lines were seen in absorption towards three target positions (SMM 5, SMM 6, and SMM 7). In SMM 6, the absorption line spectra are likely seen against the continuum emission arising from the central YSO, while towards SMM 7 an extragalactic background source could act as the source of continuum emission against which the line absorption in the clump medium is caused. However, the origin of the absorption lines in the IR dark clump SMM 5 remain unclear.

  • 3.

    The C2H molecules were found to be the most abundant ones of the detected species (~2 × 10−9 on average), while HN13C is the least abundant species (~2 × 10−11 on average).

  • 4.

    We found three potential positive correlations between the derived fractional abundances of the molecules. With a correlation coefficient of r ≃ 0.9, the most significant of these are between C2H and HN13C, and between HN13C and H13CO+. These relationships can be understood as manifestations of the gas-phase electron (ionisation degree) and atomic carbon abundances.

  • 5.

    The fractional abundances of the detected molecules in the Seahorse IRDC are typically low compared to other IRDC clumps. The only exceptions are H13CN and H13CO+, which appear to be comparably abundant (or even a few times more abundant) in the Seahorse than in other Galactic IRDCs.

  • 6.

    The statistically most signifcant evolutionary trends we found are the drop in the C2H abundance and in the [HN13C]∕[H13CN] ratio as the clump evolves from an IR dark stage to IR bright stage, and further to an H II region stage. These are in agreement with previous studies, and the former trend can be explained by the conversion of C2H to species like CO when the clump temperature rises, especially owing to the emergence of hot cores in high-mass star-forming clumps. The latter trend is likely to be a manifestation of the fairly well established negative temperature dependence of the [HNC]∕[HCN] ratio, which in turn is likely predominantly driven by neutral-neutral reactions HNC + H and HNC + O that destroy HNC (and form HCN in the first case) in the heated gas phase.

The Seahorse IRDC lies about 1. °3 or ~ 60 pc above the Galactic plane and, hence, it was not part of the Galactic regions mapped with the Spitzer IR satellite. In this sense, the Seahorse is different from the more mainstream IRDCs that have been studied in more detail and which are typically selected on the basis of their appearance in the Spitzer mid-IR images. The Seahorse’s environment and its hypothetical evolutionary stage of stillbeing in the process of accreting mass along the dusty streams, as seen in Herschel images, might have an effect on the physical and chemical properties of the cloud. Molecular line imaging of the cloud would be useful to get the picture of how the emission from different molecules is spatially distributed and to understand whether thestriations projected towards the Seahorse are physically related to the main filament. In particular, mapping the filament in a low rotational transition of SiO could help to understand if the cloud is associated with a large-scale shock front driven by its precursor colliding gas flows. On the other hand, high-resolution interferometric spectral lineimaging is required to search for HMCs within the clumps of the Seahorse IRDC and, hence, to see whether high-mass star formation is taking place elsewhere in the cloud besides the H II region associated with IRAS 13039-6108.

Acknowledgements

I would like to thank the anonymous referee for providing constructive, scientific comments that led to improvements in this paper. I am grateful to the staff at the APEX telescope for performing the service mode SEPIA observations presented inthis paper. This research has made use of NASA’s Astrophysics Data System Bibliographic Services. This research made use of Astropy (http://www.astropy.org), a community-developed core Python package for Astronomy (Astropy Collaboration 2013, 2018). I also acknowledge the community effort devoted to the development of the following open-source Python packages and libraries that were used in this work: matplotlib (Hunter 2007), pandas (McKinney 2010), numpy (van der Walt et al. 2011), seaborn (Waskom et al. 2017), and scipy (Virtanen et al. 2020). The title of this paper is inspired by the paper by Ward-Thompson et al. (2006), “SCUBA observations of the Horsehead nebula - what did the horse swallow?”.

Appendix A Spectra and line parameters

The SO(NJ = 44−33), H13CN(J = 2−1), H13CO+(J = 2−1), SiO(J = 4−3), HN13C(J = 2−1), and C2 H(N = 2−1) spectra towards all the 14 target sources are shown in Figs. A.1–A.6. The derived line parameters are given in Table A.1. The last two columns of Table A.1 also list the derived molecular column densities and fractional abundances with respect to H2.

thumbnail Fig. A.1

SO(44−33) spectra towards the Seahorse IRDC clumps. Gaussian fits to the lines are overlaid in green. While the velocity range shown in each panel is the same, the intensity range is different to better show the line profiles. The SO line is seen in absoprtion towards SMM 5–7.
Open with DEXTER
thumbnail Fig. A.2

H13CN(2−1) spectra towards the Seahorse IRDC clumps. Hyperfine structure fits to the lines are overlaid in green. While the velocity range shown in each panel is the same, the intensity range is different to better show the line profiles. The H13CN line is seen in absoprtion towards SMM 5 (and a hint of it in SMM 6). The red, vertical dashed line in the SMM 6 panel shows the systemic velocity derived from C17O(J = 2−1) by Miettinen (2012).
Open with DEXTER
thumbnail Fig. A.3

H13CO+(2−) spectra towards the Seahorse IRDC clumps. Hyperfine structure fits to the lines are overlaid in green. While the velocity range shown in each panel is the same, the intensity range is different to better show the line profiles. The H13CO+ line is seen in absoprtion towards SMM 5–7. The red, vertical dashed line in the SMM 6 panel shows the systemic velocity derived fromC17O(J = 2−1) by Miettinen (2012).
Open with DEXTER
thumbnail Fig. A.4

SiO(4−3) spectra towards the Seahorse IRDC clumps. Gaussian fits to the detected lines are overlaid in green (only one source, SMM 3). The velocity range in the SMM 3 panel is 1.8 times wider than in the other panels to better see the broad line profile. The intensity range in the panels is different to better show the line profiles.
Open with DEXTER
thumbnail Fig. A.5

HN13C(2−1) spectra towards the Seahorse IRDC clumps. Gaussian fits to the lines are overlaid in green. While the velocity range shown in each panel is the same, the intensity range is different to better show the line profiles. The HN13C line is seen in absoprtion towards SMM 5–7. The red, vertical dashed line in the SMM 6 panel shows the systemic velocity derived fromC17O(J = 2−1) by Miettinen (2012).
Open with DEXTER
thumbnail Fig. A.6

C2H(2−1) spectra towards the Seahorse IRDC clumps. The seven detected hyperfine components are labelled in the SMM 1 panel. The hyperfine structure fits to the lines are not shown to better see the detected lines. While the velocity range shown in each panel is the same (we note that it is much wider than in Figs. A.1–A.5), the intensity range is different to better show the detected lines. The C2 H lines are seen in absoprtion towards SMM 5 and 6 (also towards SMM 7 but not at our detection limit of ≥ 3σ).
Open with DEXTER
Table A.1

Spectral line parameters, molecular column densities, and fractional abundances.

References

  1. Allen, T. L., Goddard, J. D., & Schaefer, H. F. 1980, J. Chem. Phys., 73, 3255 [NASA ADS] [CrossRef] [Google Scholar]
  2. Astropy Collaboration (Robitaille, T. P., et al.) 2013, A&A, 558, A33 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  3. Astropy Collaboration (Price-Whelan, A. M., et al.) 2018, AJ, 156, 123 [NASA ADS] [CrossRef] [Google Scholar]
  4. Battersby, C., Bally, J., Jackson, J. M., et al. 2010, ApJ, 721, 222 [NASA ADS] [CrossRef] [Google Scholar]
  5. Belitsky, V., Lapkin, I., Fredrixon, M., et al. 2018, A&A, 612, A23 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  6. Beltrán, M. T., Brand, J., Cesaroni, R., et al. 2006, A&A, 447, 221 [NASA ADS] [CrossRef] [EDP Sciences] [MathSciNet] [Google Scholar]
  7. Beuther, H., & Steinacker, J. 2007, ApJ, 656, L85 [NASA ADS] [CrossRef] [Google Scholar]
  8. Beuther, H., Semenov, D., Henning, T., et al. 2008, ApJ, 675, L33 [NASA ADS] [CrossRef] [Google Scholar]
  9. Beuther, H., Ragan, S. E., Johnston, K., et al. 2015, A&A, 584, A67 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  10. Busquet, G., Estalella, R., Palau, A., et al. 2016, ApJ, 819, 139 [NASA ADS] [CrossRef] [Google Scholar]
  11. Chambers, E. T., Jackson, J. M., Rathborne, J. M., et al. 2009, ApJS, 181, 360 [NASA ADS] [CrossRef] [Google Scholar]
  12. Colzi, L., Fontani, F., Caselli, P., et al. 2018, A&A, 609, A129 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  13. Cosentino, G., Jiménez-Serra, I., Henshaw, J. D., et al. 2018, MNRAS, 474, 3760 [NASA ADS] [Google Scholar]
  14. Doty, S. D., & Neufeld, D. A. 1997, ApJ, 489, 122 [NASA ADS] [CrossRef] [Google Scholar]
  15. Dumke, M., & Mac-Auliffe, F. 2010, Proc. SPIE, 7737 77371J [CrossRef] [Google Scholar]
  16. Egan, M. P., Shipman, R. F., Price, S. D., et al. 1998, ApJ, 494, L199 [NASA ADS] [CrossRef] [Google Scholar]
  17. Favre, C., Carvajal, M., Field, D., et al. 2014, ApJS, 215, 25 [NASA ADS] [CrossRef] [Google Scholar]
  18. Finn, S. C., Jackson, J. M., Rathborne, J. M., et al. 2013, ApJ, 764, 102 [NASA ADS] [CrossRef] [Google Scholar]
  19. Fixsen, D. J. 2009, ApJ, 707, 916 [NASA ADS] [CrossRef] [Google Scholar]
  20. Frerking, M. A., Langer, W. D., & Wilson, R. W. 1979, ApJ, 232, L65 [NASA ADS] [CrossRef] [Google Scholar]
  21. Fuchs, U., Brünken, S., Fuchs, G. W., et al. 2004, Zeitschrift Naturforschung Teil A, 59, 861 [Google Scholar]
  22. Fuente, A., Martin-Pintado, J., Cernicharo, J., et al. 1993, A&A, 276, 473 [Google Scholar]
  23. Gerner, T., Beuther, H., Semenov, D., et al. 2014, A&A, 563, A97 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  24. Giannetti, A., Brand, J., Sánchez-Monge, Á., et al. 2013, A&A, 556, A16 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  25. Giannetti, A., Wyrowski, F., Leurini, S., et al. 2015, A&A, 580, L7 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  26. Godard, B., Falgarone, E., Gerin, M., et al. 2010, A&A, 520, A20 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  27. Goldsmith, P. F., Irvine, W. M., Hjalmarson, A., et al. 1986, ApJ, 310, 383 [NASA ADS] [CrossRef] [PubMed] [Google Scholar]
  28. Güsten, R., Nyman, L. Å., Schilke, P., et al. 2006, A&A, 454, L13 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  29. Hacar, A., Bosman, A. D., & van Dishoeck, E. F. 2020, A&A, 635, A4 [CrossRef] [EDP Sciences] [Google Scholar]
  30. Heitsch, F., Slyz, A. D., Devriendt, J. E. G., et al. 2006, ApJ, 648, 1052 [NASA ADS] [CrossRef] [Google Scholar]
  31. Helsel, D. R. 2005, Nondetects And Data Analysis: Statistics for Censored Environmental Data (New York: John Wiley and Sons) [Google Scholar]
  32. Henshaw, J. D., Caselli, P., Fontani, F., et al. 2013, MNRAS, 428, 3425 [NASA ADS] [CrossRef] [Google Scholar]
  33. Henshaw, J. D., Caselli, P., Fontani, F., et al. 2016, MNRAS, 463, 146 [NASA ADS] [CrossRef] [Google Scholar]
  34. Herbst, E. 1978, ApJ, 222, 508 [NASA ADS] [CrossRef] [Google Scholar]
  35. Herbst, E., & Klemperer, W. 1973, ApJ, 185, 505 [NASA ADS] [CrossRef] [Google Scholar]
  36. Herbst, E., Terzieva, R., & Talbi, D. 2000, MNRAS, 311, 869 [NASA ADS] [CrossRef] [Google Scholar]
  37. Hily-Blant, P., Walmsley, M., Pineau Des Forêts, G., et al. 2010, A&A, 513, A41 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  38. Hirota, T., Yamamoto, S., Mikami, H., et al. 1998, ApJ, 503, 717 [NASA ADS] [CrossRef] [Google Scholar]
  39. Hirota, T., Ikeda, M., & Yamamoto, S. 2003, ApJ, 594, 859 [NASA ADS] [CrossRef] [Google Scholar]
  40. Hoq, S., Clemens, D. P., Guzmán, A. E., et al. 2017, ApJ, 836, 199 [NASA ADS] [CrossRef] [Google Scholar]
  41. Hunter, J. D. 2007, Comput. Sci Eng., 9, 90 [NASA ADS] [CrossRef] [Google Scholar]
  42. Jackson, J. M., Finn, S. C., Rathborne, J. M., et al. 2008, ApJ, 680, 349 [NASA ADS] [CrossRef] [Google Scholar]
  43. Jackson, J. M., Finn, S. C., Chambers, E. T., et al. 2010, ApJ, 719, L185 [NASA ADS] [CrossRef] [Google Scholar]
  44. Jiménez-Serra, I., Caselli, P., Tan, J. C., et al. 2010, MNRAS, 406, 187 [NASA ADS] [CrossRef] [Google Scholar]
  45. Jin, M., Lee, J.-E., & Kim, K.-T. 2015, ApJS, 219, 2 [NASA ADS] [CrossRef] [Google Scholar]
  46. Juvela, M., Guillet, V., Liu, T., et al. 2018, A&A, 620, A26 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  47. Kaplan, E. L., & Meier, P. 1958, J. Am. Stat. Assoc., 53,457 [Google Scholar]
  48. Kauffmann, J., & Pillai, T. 2010, ApJ, 723, L7 [NASA ADS] [CrossRef] [Google Scholar]
  49. Kauffmann, J., Bertoldi, F., Bourke, T. L., et al. 2008, A&A, 487, 993 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  50. Langer, W. D., Graedel, T. E., Frerking, M. A., et al. 1984, ApJ, 277, 581 [NASA ADS] [CrossRef] [Google Scholar]
  51. Lee, L. 2017, NADA: Nondetects and Data Analysis for Environmental Data. R package version 1.6–1. https://CRAN.R-project.org/package=NADA [Google Scholar]
  52. Li, S., Wang, J., Fang, M., et al. 2019, ApJ, 878, 29 [NASA ADS] [CrossRef] [Google Scholar]
  53. Liu, X.-L., Wang, J.-J., & Xu, J.-L. 2013, MNRAS, 431, 27 [NASA ADS] [CrossRef] [Google Scholar]
  54. Liszt, H. S. 2017, ApJ, 835, 138 [NASA ADS] [CrossRef] [Google Scholar]
  55. Loison, J.-C., Wakelam, V., & Hickson, K. M. 2014, MNRAS, 443, 398 [NASA ADS] [CrossRef] [Google Scholar]
  56. Mangum, J. G., & Shirley, Y. L. 2015, PASP, 127, 266 [NASA ADS] [CrossRef] [Google Scholar]
  57. Maret, S., Hily-Blant, P., Pety, J., et al. 2011, A&A, 526, A47 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  58. Mattern, M., Kainulainen, J., Zhang, M., et al. 2018, A&A, 616, A78 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  59. McKee, C. F., & Krumholz, M. R. 2010, ApJ, 709, 308 [NASA ADS] [CrossRef] [Google Scholar]
  60. McKinney, W. 2010, Proceedings of the 9th Python in Science Conference, 51 [Google Scholar]
  61. Miettinen, O. 2012, A&A, 540, A104 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  62. Miettinen, O. 2014, A&A, 562, A3 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  63. Miettinen, O. 2018, A&A, 609, A123 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  64. Miettinen, O., & Harju, J. 2010, A&A, 520, A102 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  65. Millar, T. J. 2015, Plasma Sources Sci. Technol., 24, 043001 [NASA ADS] [CrossRef] [Google Scholar]
  66. Mladenović, M., & Roueff, E. 2017, A&A, 605, A22 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  67. Moore, D. S., Notz, W. I., & Fligner, M. A. 2018, The Basic Practice of Statistics, 8th edn. (New York, NY: W. H. Freeman and Company) [Google Scholar]
  68. Motte, F., Bontemps, S., & Louvet, F. 2018, ARA&A, 56, 41 [NASA ADS] [CrossRef] [Google Scholar]
  69. Müller, H. S. P., Schlöder, F., Stutzki, J., et al. 2005, J. Mol. Struct., 742, 215 [NASA ADS] [CrossRef] [Google Scholar]
  70. Nagy, Z., Ossenkopf, V., van der Tak, F. F. S., et al. 2015, A&A, 578, A124 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  71. Ossenkopf, V.,& Henning, T. 1994, A&A, 291, 943 [Google Scholar]
  72. Padoan, P., Pan, L., Juvela, M., et al. 2019, ApJ, submitted [arXiv:1911.04465] [Google Scholar]
  73. Padovani, M., Walmsley, C. M., Tafalla, M., et al. 2009, A&A, 505, 1199 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  74. Padovani, M., Walmsley, C. M., Tafalla, M., et al. 2011, A&A, 534, A77 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  75. Pearson, P. K., & Schaefer, H. F. 1974, ApJ, 192, 33 [NASA ADS] [CrossRef] [Google Scholar]
  76. Pérault, M., Omont, A., Simon, G., et al. 1996, A&A, 315, L165 [Google Scholar]
  77. Peretto, N., & Fuller, G. A. 2009, A&A, 505, 405 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  78. Pickett, H. M., Poynter, R. L., Cohen, E. A., et al. 1998, J. Quant. Spectr. Rad. Transf., 60, 883 [NASA ADS] [CrossRef] [Google Scholar]
  79. Pilbratt, G. L., Riedinger, J. R., Passvogel, T., et al. 2010, A&A, 518, L1 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  80. Pillai, T., Wyrowski, F., Menten, K. M., et al. 2006, A&A, 447, 929 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  81. R Core Team 2019 R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria, http://www.R-project.org/ [Google Scholar]
  82. Ragan, S. E., Henning, T., & Beuther, H. 2013, A&A, 559, A79 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  83. Rathborne, J. M., Jackson, J. M., Chambers, E. T., et al. 2005, ApJ, 630, L181 [NASA ADS] [CrossRef] [Google Scholar]
  84. Rathborne, J. M., Jackson, J. M., & Simon, R. 2006, ApJ, 641, 389 [NASA ADS] [CrossRef] [Google Scholar]
  85. Rathborne, J. M., Jackson, J. M., Chambers, E. T., et al. 2010, ApJ, 715, 310 [NASA ADS] [CrossRef] [Google Scholar]
  86. Rathborne, J. M., Garay, G., Jackson, J. M., et al. 2011, ApJ, 741, 120 [NASA ADS] [CrossRef] [Google Scholar]
  87. Rawlings, J. M. C., Redman, M. P., Keto, E., et al. 2004, MNRAS, 351, 1054 [NASA ADS] [CrossRef] [Google Scholar]
  88. Reitblat, A. A. 1980, Sov. Astron. Lett., 6, 406 [NASA ADS] [Google Scholar]
  89. Roberts, J. F., Jiménez-Serra, I., Gusdorf, A., et al. 2012, A&A, 544, A150 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  90. Ruze, J. 1952, Il Nuovo Cimento, 9, 364 [NASA ADS] [CrossRef] [Google Scholar]
  91. Sakai, T., Sakai, N., Hirota, T., et al. 2010, ApJ, 714, 1658 [NASA ADS] [CrossRef] [Google Scholar]
  92. Sakai, T., Sakai, N., Furuya, K., et al. 2012, ApJ, 747, 140 [NASA ADS] [CrossRef] [Google Scholar]
  93. Sánchez-Monge, Á., Beltrán, M. T., Cesaroni, R., et al. 2013, A&A, 550, A21 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  94. Sanhueza, P., Jackson, J. M., Foster, J. B., et al. 2012, ApJ, 756, 60 [NASA ADS] [CrossRef] [Google Scholar]
  95. Saral, G., Audard, M., & Wang, Y. 2018, A&A, 620, A158 [CrossRef] [EDP Sciences] [Google Scholar]
  96. Sarrasin, E., Abdallah, D. B., Wernli, M., et al. 2010, MNRAS, 404, 518 [Google Scholar]
  97. Schilke, P., Walmsley, C. M., Pineau Des Forets, G., et al. 1992, A&A, 256, 595 [Google Scholar]
  98. Schmid-Burgk, J., Muders, D., Müller, H. S. P., et al. 2004, A&A, 419, 949 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  99. Schuller, F., Menten, K. M., Contreras, Y., et al. 2009, A&A, 504, 415 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  100. Simon, R., Jackson, J. M., Rathborne, J. M., et al. 2006, ApJ, 639, 227 [NASA ADS] [CrossRef] [Google Scholar]
  101. Siringo, G., Kreysa, E., Kovács, A., et al. 2009, A&A, 497, 945 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  102. Soam, A., Liu,T., Andersson, B.-G., et al. 2019, ApJ, 883, 95 [CrossRef] [Google Scholar]
  103. Svoboda, B. E., Shirley, Y. L., Traficante, A., et al. 2019, ApJ, 886, 36 [NASA ADS] [CrossRef] [Google Scholar]
  104. Tang, Y.-W., Koch, P. M., Peretto, N., et al. 2019, ApJ, 878, 10 [CrossRef] [Google Scholar]
  105. Taniguchi, K., Saito, M., Sridharan, T. K., et al. 2019, ApJ, 872, 154 [CrossRef] [Google Scholar]
  106. Turner, B. E. 1991, ApJS, 76, 617 [NASA ADS] [CrossRef] [Google Scholar]
  107. Turner, B. E. 2001, ApJS, 136, 579 [NASA ADS] [CrossRef] [Google Scholar]
  108. Turner, B. E., Herbst, E., & Terzieva, R. 2000, ApJS, 126, 427 [NASA ADS] [CrossRef] [Google Scholar]
  109. van der Tak, F. F. S., Müller, H. S. P., Harding, M. E., et al. 2009, A&A, 507, 347 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  110. van der Walt, S., Colbert, S. C., & Varoquaux, G. 2011, Comput. Sci. Eng., 13, 22 [Google Scholar]
  111. Vasyunina, T., Linz, H., Henning, T., et al. 2009, A&A, 499, 149 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  112. Vasyunina, T., Linz, H., Henning, T., et al. 2011, A&A, 527, A88 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  113. Vasyunina, T., Vasyunin, A. I., Herbst, E., et al. 2014, ApJ, 780, 85 [NASA ADS] [CrossRef] [Google Scholar]
  114. Virtanen, P., Gommers, R., Oliphant, T. E., et al. 2020, Nature Methods, 17, 261 [CrossRef] [Google Scholar]
  115. Visser, R., vanDishoeck, E. F., & Black, J. H. 2009, A&A, 503, 323 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  116. Walsh, A. J., Thorwirth, S., Beuther, H., et al. 2010, MNRAS, 404, 1396 [Google Scholar]
  117. Wang, Y., Zhang, Q., Rathborne, J. M., et al. 2006, ApJ, 651, L125 [NASA ADS] [CrossRef] [Google Scholar]
  118. Wang, K., Zhang, Q., Wu, Y., et al. 2011, ApJ, 735, 64 [NASA ADS] [CrossRef] [Google Scholar]
  119. Ward-Thompson, D., Nutter, D., Bontemps, S., et al. 2006, MNRAS, 369, 1201 [NASA ADS] [CrossRef] [Google Scholar]
  120. Waskom, M., Botvinnik, O., O’Kane, D., et al. 2017, mwaskom/seaborn: v0.8.1, available from https://doi.org/10.5281/zenodo.883859 [Google Scholar]
  121. Watt, G. D., White, G. J., Millar, T. J., et al. 1988, A&A, 195, 257 [Google Scholar]
  122. Wright, E. L., Eisenhardt, P. R. M., Mainzer, A. K., et al. 2010, AJ, 140, 1868 [Google Scholar]
  123. Zeng, S., Jiménez-Serra, I., Cosentino, G., et al. 2017, A&A, 603, A22 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
1

Herschel is an ESA space observatory with science instruments pro- vided by European-led Principal Investigator consortia and with important participation from NASA.

3

The APEX atmospheric transmission calculator is available at http://www.apex-telescope.org/sites/chajnantor/atmosphere/transpwv/

4

The APEX antenna efficiences can be found at http://www.apex-telescope.org/telescope/efficiency/index.php

5

Grenoble Image and Line Data Analysis Software (GILDAS) is provided and actively developed by Institut de Radioastronomie Millimétrique (IRAM), and is available at http://www.iram.fr/IRAMFR/GILDAS

All Tables

Table 1

Source sample.

In the text
Table 2

Observational parameters.

In the text
Table 3

Detected spectral lines and selected spectroscopic parameters.

In the text
Table 4

Statistics of the molecular abundances.

In the text
Table 5

HN13C/H13CN and HN13C/H13CO+ abundance ratios.

In the text
Table 6

Equations of the linear regressions plotted in Fig. 4 and the corresponding correlation coefficients.

In the text
Table A.1

Spectral line parameters, molecular column densities, and fractional abundances.

In the text

All Figures

thumbnail Fig. 1

Herschel/SPIRE (Spectral and Photometric Imaging REceiver) 350 μm image towards the IRDC G304.74+01.32 (the Seahorse IRDC) and its surroundings. The image is displayed using a non-linear (logarithmic) stretch to improve the visibility of the faint, extended dust emission. There appear to be dusty striations on the western side of the Seahorse filament. A scale bar of 3 pc projected length is shown in the bottom left corner.
Open with DEXTER
In the text
thumbnail Fig. 2

WISE 12 μm image towards the IRDC G304.74+01.32 (the Seahorse IRDC). The image is displayed using a logarithmic stretch to improve the colour contrast. The overlaid contours represent the LABOCA 870 μm emission (Miettinen & Harju 2010; Miettinen 2018); the contours start at 3σ, and increase in steps of 3σ, where 3σ = 120 mJy beam−1. The target clumps are labelled so that the numbers refer to the SMM IDs (e.g. 1 refers to SMM 1), while the sources I13037, I13039, and I13042 are the three IRAS sources in the filament. The sources BLOB 1 and BLOB 2 are labelled as B1 and B2. The red plus signs indicate the target positions of the present spectral line observations (i.e. LABOCA 870 μm emissionpeaks of the clumps). A scale bar of 1 pc projected length, and the beam size at the tuned frequency of the present spectral line observations (35.′′ 9 HPBW) are shown in the bottom left corner.
Open with DEXTER
In the text
thumbnail Fig. 3

Box plot (or box-and-whisker plot) showing the distributions of the derived molecular fractional abundances. The boxes are arranged in ascending order according to the median value. The height of the box represents the interquartile range IQR ≡ [Q1, Q3] (i.e. from the 25th to 75th percentile). A horizontal line that goes through the box shows the median abundance (Q2 or the 50th percentile). The whiskers are defined to extend from Q1 − 1.5 ×IQR to Q3 +1.5 ×IQR unless the minimum and maximum values of the distribution are encountered before that (in which case the whiskers extend to the minimum and maximum values). We note that only the uncensored data are used to plot the distributions (i.e. the upper limits are ignored), and SiO was detected in only one of the target sources (SMM 3).
Open with DEXTER
In the text
thumbnail Fig. 4

Scatter plots between the variables that were found to be linearly correlated with a Pearson correlation coefficient of r > 0.7. The arrows pointing down indicate upper limits. The dashed lines show the linear regression fits to the uncensored data points (i.e. the upper limits were not taken into account).
Open with DEXTER
In the text
thumbnail Fig. 5

Zoom-in of Fig. 2 towards SMM 5, 6, and 7 for which some of the spectral lines are seen in absorption. Only the 3σ and 6σ LABOCA contours are plotted for clarity. Besides the clumps SMM 5, 6, and 7, also the striation seen as a dark absorption feature is labelled (cf. Fig. 1, where the striation is seen in dust emission). The red plus signs indicate the target positions of the present spectral line observations, while the yellow diamond symbols show the positions of the WISE sources seen towards SMM 6 and SMM 7. The latter one is likely to be an extragalactic object, thatis a star-forming galaxy or an active galactic nucleus, and hence SMM 7 is considered an IR dark clump (Miettinen 2018). A scale bar of 0.5 pc projected length, and the beam size at the tuning frequency of the present spectral lineobservations are shown in the bottom left corner.
Open with DEXTER
In the text
thumbnail Fig. 6

Comparison of the molecular abundance distributions (column density for SO) in the Seahorse IRDC with those in other IRDC sources presented in the literature. The box plots are sorted by increasing median abundance. The data points above some of the boxes represent outlier values. Upper limits are neglected from the distributions.
Open with DEXTER
In the text
thumbnail Fig. 7

Box plots showing the clump properties that were found to be different on average for the different types of clumps in the present sample, namely clumps that are either dark or bright in the mid-IR and clumps associated with an H II region (onlyone source, IRAS 13039-6108). Censored data are not included in the plot. Some of the trends could be manifestations of clump evolution (Sect. 4.4).
Open with DEXTER
In the text
thumbnail Fig. A.1

SO(44−33) spectra towards the Seahorse IRDC clumps. Gaussian fits to the lines are overlaid in green. While the velocity range shown in each panel is the same, the intensity range is different to better show the line profiles. The SO line is seen in absoprtion towards SMM 5–7.
Open with DEXTER
In the text
thumbnail Fig. A.2

H13CN(2−1) spectra towards the Seahorse IRDC clumps. Hyperfine structure fits to the lines are overlaid in green. While the velocity range shown in each panel is the same, the intensity range is different to better show the line profiles. The H13CN line is seen in absoprtion towards SMM 5 (and a hint of it in SMM 6). The red, vertical dashed line in the SMM 6 panel shows the systemic velocity derived from C17O(J = 2−1) by Miettinen (2012).
Open with DEXTER
In the text
thumbnail Fig. A.3

H13CO+(2−) spectra towards the Seahorse IRDC clumps. Hyperfine structure fits to the lines are overlaid in green. While the velocity range shown in each panel is the same, the intensity range is different to better show the line profiles. The H13CO+ line is seen in absoprtion towards SMM 5–7. The red, vertical dashed line in the SMM 6 panel shows the systemic velocity derived fromC17O(J = 2−1) by Miettinen (2012).
Open with DEXTER
In the text
thumbnail Fig. A.4

SiO(4−3) spectra towards the Seahorse IRDC clumps. Gaussian fits to the detected lines are overlaid in green (only one source, SMM 3). The velocity range in the SMM 3 panel is 1.8 times wider than in the other panels to better see the broad line profile. The intensity range in the panels is different to better show the line profiles.
Open with DEXTER
In the text
thumbnail Fig. A.5

HN13C(2−1) spectra towards the Seahorse IRDC clumps. Gaussian fits to the lines are overlaid in green. While the velocity range shown in each panel is the same, the intensity range is different to better show the line profiles. The HN13C line is seen in absoprtion towards SMM 5–7. The red, vertical dashed line in the SMM 6 panel shows the systemic velocity derived fromC17O(J = 2−1) by Miettinen (2012).
Open with DEXTER
In the text
thumbnail Fig. A.6

C2H(2−1) spectra towards the Seahorse IRDC clumps. The seven detected hyperfine components are labelled in the SMM 1 panel. The hyperfine structure fits to the lines are not shown to better see the detected lines. While the velocity range shown in each panel is the same (we note that it is much wider than in Figs. A.1–A.5), the intensity range is different to better show the detected lines. The C2 H lines are seen in absoprtion towards SMM 5 and 6 (also towards SMM 7 but not at our detection limit of ≥ 3σ).
Open with DEXTER
In the text

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.