EDP Sciences
Free Access
Issue
A&A
Volume 602, June 2017
Article Number L3
Number of page(s) 6
Section Letters
DOI https://doi.org/10.1051/0004-6361/201629792
Published online 31 May 2017

© ESO, 2017

1. Introduction

In molecular cloud cores, the formation of deuterated molecules is favoured at low temperatures (T ≤ 20 K) and at high densities (n ≥ 104 cm-3). It is thus expected that the deuteration fraction (Dfrac, the relative abundance between a species containing D as compared to the same species containing H) is enhanced in cold and dense prestellar cores. Dfrac should then decrease after protostellar birth, when the young stellar object heats the central region of the core (Caselli 2002). For low-mass star-forming cores, observations of deuterated species produced in gas-phase reactions, such as H2D+ and N2D+, have confirmed this theoretical scenario (Crapsi et al. 2005; Caselli et al. 2008).

Recent studies show that high Dfrac values are also typical for high-mass star-forming cores and that the Dfrac of some species could also be an evolutionary indicator in the intermediate- and high-mass regime (e.g. Busquet et al. 2010; Fontani et al. 2011; Sakai et al. 2012). High-mass star-forming regions can be divided into the following evolutionary stages: high-mass starless cores (HMSCs), high-mass protostellar objects (HMPOs), and ultracompact HII regions (UCHIIs; e.g. Beuther et al. 2007; Tan et al. 2014). By studying several deuterated species in 27 massive cores, Fontani et al. (2011, 2015) found that species formed exclusively in the gas (N2H+) showed different evolutionary trends from those formed partially (NH3) or totally (CH3OH) on grain mantles. The abundance of N2D+ is indeed higher in HMSCs, and it drops by about an order of magnitude during the HMPO and UCHII stages. This is due to the higher gas temperatures found in the latter objects, which allow the destruction of H2D+ via the endothermic reaction H2D+ + H2 H + HD (see Gerlich et al. 2002; Fontani et al. 2011). In contrast, deuterated methanol, formed only on grain surfaces, is detected towards HMPOs and externally heated HMSCs only, possibly as a result of evaporation or sputtering of grain mantles. Therefore, while Dfrac(N2H+) can be used as an indicator of the initial conditions in starless or pre-stellar cores, high values of Dfrac(CH3OH) are a good probe of the earliest protostellar phases (Fontani et al. 2015). The deuteration fraction of NH3, whose formation pathways occur both in the gas-phase and on grain surfaces, does not show statistically significant changes with evolution, and therefore there is no dominant formation pathway for NH3 and its deuterated forms.

Like NH3, formaldehyde (H2CO) may also be produced both in the gas phase and on grain surfaces. The two main pathways for the production of H2CO (and its deuterated counterparts) involve CH (viz. CH2D+ and CHD) in the gas phase and multiple hydrogenation (viz. deuteration) of CO in the ices (see e.g. Roberts & Millar 2007). The gas phase pathway is similar to the one of N2H+ with the main difference that deuteration of H2CO can also occur at warmer temperatures (3050 K, see Parise et al. 2009), whereas T< 20 K are needed to increase the deuteration of N2H+ via H2D+. The ice phase formation route of H2CO is similar to that of CH3OH and their deuterated forms. Although laboratory work and observational studies of low-mass protostars suggests an important contribution from grain surface chemistry for the production of H2CO (Watanabe 2005; Roberts & Millar 2007; Bergman et al. 2011), the relative importance of the dust grain vs. gas-phase formation routes remains unclear. Furthermore, D2CO measurements in intermediate- and in high-mass star-forming regions are lacking (D2CO has been firmly detected toward NGC 7129-FIRS 2 and tentatively toward the MonR2 ultra-compact HII region; Fuente et al. 2005; Treviño-Morales et al. 2014), and it is thus unknown whether the chemistry of H2CO and of its deuterated counterparts is governed by the same mechanisms in both low-mass and more massive objects.

In this Letter, we report the first detection of doubly deuterated formaldehyde, D2CO, toward high-mass star-forming cores. These are the initial findings of a search for HDCO and D2CO emission toward three high-mass star-forming regions at different evolutionary stages using the new SEPIA Band 5 receiver available at the Atacama Pathfinder EXperiment (APEX1) telescope. The targets were extracted from the sample of Fontani et al. (2011), for which the deuteration fraction of other molecules (e.g. N2H+, CH3OH, and NH3) has been measured (Fontani et al. 2015). The evolutionary stages of our sample reach from the initial conditions in HMSCs to the HMPOs phase and the UCHII regions. The source sample and the observations are described in Sect. 2, while our results are presented in Sect. 3. In Sect. 4 we discuss our results and place them in context with respect to previous findings in low-mass star-forming regions. In Sect. 5 we summarize our conclusions.

2. Observations

2.1. Selected targets

The selected sources are extracted from Fontani et al. (2011) and are AFGL5142, IRAS 05358+3543, and G5.89-0.39. These are the sources with the brightest lines of H2CO within each evolutionary stage. We note that this selection may introduce biases, therefore a follow-up study with a larger sample is needed. The bolometric luminosities of the objects are 103.6, 103.8, and 105.1L and the integrated gas masses are 210, 300, and 300 M for AFGL5142, IRAS 05358, and G5.89, respectively (Fontani et al. 2011; Liu et al. 2016; Beuther et al. 2002b, 2005; Tang et al. 2009). AFGL5142 is a high-mass star-forming region (distance of 2.14 ± 0.05 kpc; Burns et al., in prep.) that hosts several dense cores at different evolutionary stages: the central core (CC, classified as an HMPO), the western core (WC, an HMSC core), and the eastern core (EC; see Busquet et al. 2011; Fontani et al. 2011). We centred our single-pointing observations on CC. However, because the separation between WC and CC is 9′′ and because the APEX beam is ~34 at 190 GHz, WC and CC were also covered in our observations. The N2H+ (10) emission in AFGL5142 shows two velocity components at 2 and 4 km s-1, associated with CC and WC, respectively (Busquet et al. 2011).

IRAS 05358+3543, located at 1.8 kpc (Snell et al. 1990), contains three dust condensations (mm1, mm2, and mm3) within an area of 9″× 4″ (Beuther et al. 2002a). According to Leurini et al. (2007), mm1 splits into mm1a, a hot core and a massive circumstellar disk with T ~ 220 K, and mm1b, which is at an earlier stage of evolution. Source mm2 is a low-to-intermediate mass protostar, while mm3 is an HMSC. Our observation is centred on mm1, but all three dust condensations are covered within the APEX beam.

G5.89-0.39 is a shell-like UCHII region (diameter of ~4) found at 2.99 kpc (Sato et al. 2014). Sub-arcsecond observations reveal at least five dust condensations (Su et al. 2009). We centred our observations on the SMA-N dust condensation. Several energetic outflows and maser activity have been detected toward G5.89-0.39 (see e.g. Hunter et al. 2008; Fish et al. 2005).

2.2. IRAM-30 m H2CO observations

Previous observations of these objects were performed using the IRAM-30 m telescope (details can be found in Fontani et al. 2011, 2015). The frequency setups included the 303−202, 322−221, and 321−220 transitions of H2CO. The spectroscopic information of these transitions is shown in Table A.1 (Müller et al. 2005). The H2CO line emission is very strong in all three sources, with a measured TMB> 5 K (top panels in Fig. A.1).

Table 1

Calculated column densities and ratios (deuteration fractions) for the three species.

2.3. APEX SEPIA observations

The three sources were observed with the APEX SEPIA receiver (Swedish-ESO PI receiver for APEX; Billade et al. 2012)2. We carried out single-pointing observations of HDCO and D2CO using the position-switching observing mode. The J2000 central coordinates used in our observations were [RA, Dec] = (05:30:48.0, +33:47:54) for AFGL5142, [RA, Dec] = (05:39:13.0, +35:45:51) for IRAS05358+3543, and [RA, Dec] = (18:00:30.5, 24:04:01) for G5.89-0.39. The observed transitions of HDCO (at 185 GHz and 193 GHz) and D2CO (at 174.4 GHz) are shown in Table A.1. The pointing was checked every 6090 min, and the typical system temperatures were 150 K. The beam size at 183 GHz was ~34 and a beam efficiency of 0.83 was used. The XFFTS spectrometer provided a velocity resolution of 0.059 km s-1 and 0.066 km s-1 for the HDCO (at 185193 GHz) and D2CO transitions (at 174 GHz), respectively. However, for the data analysis, we smoothed the spectra to a uniform velocity resolution of 0.5 km s-1. At this velocity resolution, we reached an rms of ~0.01 K with typical integration times between 32 and 48 min. The PWV was 0.31.5 mm. The data were reduced and analysed with the Gildas software3 (version jul14a, Pety 2005) and the derived parameters of the detected lines are reported in Appendix A.

3. Results

In Fig. A.1 we present the HDCO and D2CO lines observed toward AFGL5142, IRAS 05358+3543, and G5.89-0.39. The H2CO 30,3−20,2 data obtained with the IRAM 30 m observation are also shown for comparison. While the HDCO lines (upper energy levels of 18.550.4 K) are clearly detected in all sources with peak intensities 35 mK, the D2CO 30,3−20,2 transition is only observed toward AFGL5142 and IRAS 05358+3543, that is, at the earlier HMSC/HMPO stages. The 3σ upper limit measured for the D2CO 32,2−22,1 line toward G5.89-0.39 is 6 mK.

In Table A.2 we report the derived parameters of the observed lines of H2CO, HDCO, and D2CO. For AFGL5142, the D2CO 30,3−20,2 transition shows two velocity components at −2.077 km s-1 and −3.658 km s-1 (see Table A.2). We thus fitted the entire observed H2CO, HDCO, and D2CO line emission using two Gaussian line profiles with fixed central velocities at −2.077 km s-1 and −3.658 km s-1. These two velocity components are associated with the WC core at the HMSC stage (radial velocity ~−3.5 km s-1), and with the CC core at the HMPO stage (at ~−2 km s-1; Busquet et al. 2011). Both cores lie within the single-dish beam of our IRAM 30 m and APEX observations (WC and CC are separated by 9′′, while the IRAM 30 m and APEX beam sizes are 11′′ and 34′′, respectively).

Since the H2CO, HDCO, and D2CO lines toward G5.89-0.39 show a slight asymmetry in their line profiles, we also fitted these data using two Gaussian line profiles centred at 8.5 and 11.5 km s-1. These two components can also be seen in the emission from N2H+ or NH3, for example (see Fontani et al. 2011, 2015).

For IRAS 05358+3543, H2CO, HDCO and D2CO show only one velocity component that peaks at 16 km s-1.

To estimate the molecular column densities of H2CO, HDCO, and D2CO, we used the MADCUBAIJ software that assumes LTE conditions (Martín et al. 2011; Rivilla et al. 2016). We adopted a source size of as measured from NH3(2, 2) and following Fontani et al. (2015). This source size is consistent with the size obtained in other high-density tracers such as N2H+(1-0) and C34S(7-6) (see Hunter et al. 2008; Busquet et al. 2011). We have attempted to derive the excitation temperature of the gas, Tex, using the four APEX HDCO transitions (Table A.1). Most of the observed J = 3 → 2 lines of H2CO, HDCO, and D2CO most likely trace gas with similar physical conditions given that their Eup and Aij are similar. Hence, the determination of Tex is highly uncertain because only few lines are available. We thus assumed a Tex of 28 K, which is the average temperature obtained from NH3 for IRAS 05358+3545 and AFGL5142 (Fontani et al. 2011)4. We note that the assumed value of Tex can change the derived column densities, but not their ratios. When we assume a Tex value in the 2040 K range, the variation of the column density values compared to the Tex = 28 K case is below 50%. Lower and higher excitation temperatures cannot fit the observed transitions properly. We also estimated the column densities with different source sizes. With a source size of 2′′ we cannot fit the observed transitions properly because of optical depth effects. The estimated column densities decrease by factors of 23 for a source size of 10”. For AFGL5142, we obtained the H2CO, HDCO, and D2CO column densities for the velocity components at −2.077 km s-1 (CC, in the HMPO stage) and at −3.658 km s-1 (WC in the HMSC phase). In Table 1 we report the calculated column densities of H2CO, HDCO, and D2CO and the deuteration fractions obtained from the column density ratios HDCO/H2CO, D2CO/H2CO, and D2CO/HDCO.

The HCO 31,3−21,2 transition was covered within our APEX SEPIA setup. We have estimated the H2CO column densities from the HCO column density and using the 12C/13C ratio derived by Milam et al. (2005) as a function of Galactocentric distance. The H2CO lines in AFGL5142 and G5.89-0.39 are not very optically thick since the H2CO column densities calculated using the main isotopologue H2CO and the 13C isotopologue HCO differ only by factors 0.4 and 3.6. For IRAS 05358+3543, however, the column densities derived from H2CO and HCO differ by a factor of 9. In this case, we use the column density derived directly from the H2CO lines and consider it as a lower limit for IRAS 05358.

4. Discussion

As proposed by Fontani et al. (2011, 2015), the different trends observed for Dfrac(N2H+), Dfrac(CH3OH), and Dfrac(NH3) as a function of evolution in high-mass star-forming regions are likely due to the way deuteration occurs for the different species: in the gas phase for N2H+, on the grain surface for CH3OH, and via a mixture of the two for NH3. For H2CO, D2CO is expected to be a product of material processed on solid ices since its formation timescales are longer than the depletion timescales in the pre-stellar phase (see Taquet et al. 2012). Consequently, D2CO should show an evolutionary trend similar to that derived for Dfrac(CH3OH), which peaks at the HMPO phase (Fontani et al. 2015). If HDCO were to show a similar behaviour to D2CO, we could conclude that HDCO also forms mostly on dust grains. This hypothesis would be supported by the fact that HDCO is detected in hot cores and hot corinos, while N2D+ (only formed in the gas phase) is not (see e.g. Fuente et al. 2005).

Although our sample and statistics are limited, our results show a trend for the deuteration fractions HDCO/H2CO and D2CO/H2CO to progressively increase from the HMSC to the HMPO stage by factors ~310, and to subsequently decrease at the UCHII phase (by factors 20; see Fig. 1). This behaviour is similar to the trend observed for Dfrac(CH3OH), and since the measured HDCO/H2CO are similar and D2CO/H2CO ratios are 421 times higher than those predicted by Roberts & Millar (2007) for the production of HDCO and D2CO via gas-phase reactions, surface chemistry very likely plays an important role in the formation of HDCO and D2CO.

thumbnail Fig. 1

Comparison of the average deuterated fraction of H2CO, HDCO, CH3OH, NH3, and N2H+. The upper limit for G5.89-0.39 is not shown in the left middle panel, the value is <0.001. Average values for CH3OH, NH3, and N2H+ based on Fontani et al. (2015) are shown.

Open with DEXTER

Rodgers & Charnley (2002) indeed proposed a grain surface formation scenario for these molecules in which H2CO and its deuterated forms represent intermediate steps in the formation of CH3OH and deuterated-CH3OH via hydrogenation and D-addition reactions. The extent to which surface chemistry is responsible for the production of HDCO and D2CO in these sources can be estimated by using the F parameter, (1)From the HDCO/H2CO and D2CO/H2CO ratios in Table 1, we find that F ~ 0.1−0.5 for HMSCs, F ~ 1.8 for HMPOs and F ≥ 0.2 for UCHIIs. As explained by Rodgers & Charnley (2002), low F values are consistent with grain surface chemistry. However, F values in the range 1.62.3 could also be explained by gas-phase chemistry. At the HMPO stage we therefore cannot rule out a contribution to the HDCO/D2CO production via the gas phase neutral-neutral reactions CH2D + O HDCO + H and CHD2 + O D2CO + H. Note that these reactions occur in the forward direction at temperatures ~3050 K because the formation reactions of CH2D and CHD2 are exothermic (Turner 1990).

Table 1 also shows that the D2CO/HDCO ratio remains constant at ~0.2 within the uncertainties, regardless of the source or evolutionary stage. As for the low-mass regime, the ratio HDCO/D2CO lies well below the statistical value given as D-species/D2-species = 4 × (D-species/H-species)-1, consistent with the grain surface formation scenario (see Ceccarelli et al. 2014).

When we compare the HDCO/H2CO and D2CO/H2CO ratios from Table 1 with those collected by Ceccarelli et al. (2014) for low-mass star-forming regions (see their Fig. 5), we find that the ratio in the high-mass regime, while the same ratio is close to 1 in low-mass star-forming regions. This may be a consequence of the shorter timescales available in the high-mass regime for the formation of D2CO in the ices (Taquet et al. 2012; Rodgers & Charnley 2002). In the scheme of Rodgers & Charnley (2002), D2CO is formed after HDCO through two hydrogen-deuterium exchanges experienced by H2CO in the ices. The higher ratio in high-mass star-forming regions therefore suggests that HDCO may not have had enough time to have been converted into D2CO as a consequence of the faster evolution of high-mass protostars as compared to their low-mass counterparts.

5. Summary

We have observed HDCO and D2CO toward a sample of high-mass star-forming regions at different evolutionary stages. HDCO transitions were detected for all of them, while the D2CO line was detected only for the earlier HMSC and HMPO stages. Our results point toward the idea that H2CO and its deuterated species form mostly on grain surfaces, although some gas-phase contribution is expected at the warm HMPO stage. Interferometric observations are needed to separate the HDCO and D2CO emission originating from the small and dense cores and to distinguish their origin in high-mass star-forming regions.


1

This publication is based on data acquired with the Atacama Pathfinder Experiment (APEX). APEX is a collaboration between the Max-Planck-Institut für Radioastronomie, the European Southern Observatory, and the Onsala Space Observatory.

2

Observations were done as part of Science Verification in July 2015 (project E-095.F-9808A), and in NovemberOctober 2015 and MayJune 2016 within projects E-096.C-0484A and E097.C-0897A.

4

Fontani et al. (2011) do not report any temperature for G5.89-0.39 and therefore we assume the same Tex of 28 K for this source.

Acknowledgments

We would like to thank the SEPIA team and APEX team for the successful commissioning of the instrument and for the observations. This work was partly supported by the Gothenburg Centre of Advanced Studies in Science and Technology through the program Origins of habitable planets and by the Italian Ministero dell’Istruzione, Università e Ricerca through the grant Progetti Premiali 2012 – iALMA (CUP C52I13000140001) and by NAOJ ALMA Scientific Research Grant Number 2016-03B. K.W. acknowledges the support from Deutsche Forschungsgemeinschaft (DFG) grant WA3628-1/1 through priority programme 1573 (“Physics of the Interstellar Medium”). I.J.-S. acknowledges the financial support received from the STFC through an Ernest Rutherford Fellowship (proposal number ST/L004801/1). L.V.T. and S.Z. acknowledge the support by the OTKA grants NN-111016 and K101393.

References

Appendix A: Observed H2CO, HDCO, and D2CO lines

In Table A.1 we report the parameters of the observed lines for H2CO, HDCO, and D2CO. The observing setup (and the observations of Fontani et al. 2011, 2015) also covered other higher excitation lines of the same molecules, but these were not detected (and are not expected to be detectable under the expected excitation conditions).

In Fig. A.1 we present the HDCO and D2CO lines observed toward AFGL5142, IRAS 05358+3543, and G5.89-0.39. The H2CO 30,3−20,2 data obtained with the IRAM 30 m observation are also shown for comparison. While the HDCO lines (upper energy levels of 18.5-50.4 K) are clearly detected in all sources with peak intensities 35 mK, the D2CO 30,3−20,2 transition is only observed toward AFGL5142 and IRAS 05358+3543, that is to say, at the earlier HMSC/HMPO stages. The 3σ upper limit measured for the D2CO 32,2−22,1 line toward G5.89-0.39 is 6 mK.

In Table A.2 we report the parameters derived by fitting single or multiple (in the case of AFGL5142 and G5.89-0.39) Gaussian profiles to the detected lines. For the HDCO, HCO, and H2CO spectra of AFGL5142 and IRAS 05358+3543, we fitted the vLSR and FWHM values based on the D2CO velocity components. For G5.89-0.39, we derived the velocity and width of the two components based on the HDCO 30,320,2 line. We did not use these fits directly to calculate the physical parameters. We used the velocities and line widths as initial parameters for the MADCUBAIJ modelling.

thumbnail Fig. A.1

From left to right: spectra of H2CO 30,3−20,2, HCO 31,3−21,2, HDCO 32,1−22,0, 32,2−22,1, 31,3−21,2, 30,3−20,2 and D2CO 30,3−20,2 observed toward AFGL5142, IRAS 05358+3543, and G5.89-0.39 using the IRAM 30 m telescope (Fontani et al. 2011, 2015) and the APEX SEPIA Band 5 receiver (this work). Red and blue lines indicate the fitted different velocity components for AFGL5142 and G5.89-0.39 with fixed velocity and width. The C13CS N = 15−14 and J = 15−14 transition is shown in the fourth panels from the top.

Open with DEXTER

Table A.1

Transitions of H2CO, HCO, HDCO, and D2CO observed with APEX SEPIA and the IRAM-30 m telescope.

Table A.2

Fitted parameters of the observed H2CO, HCO, HDCO, and D2CO lines.

All Tables

Table 1

Calculated column densities and ratios (deuteration fractions) for the three species.

Table A.1

Transitions of H2CO, HCO, HDCO, and D2CO observed with APEX SEPIA and the IRAM-30 m telescope.

Table A.2

Fitted parameters of the observed H2CO, HCO, HDCO, and D2CO lines.

All Figures

thumbnail Fig. 1

Comparison of the average deuterated fraction of H2CO, HDCO, CH3OH, NH3, and N2H+. The upper limit for G5.89-0.39 is not shown in the left middle panel, the value is <0.001. Average values for CH3OH, NH3, and N2H+ based on Fontani et al. (2015) are shown.

Open with DEXTER
In the text
thumbnail Fig. A.1

From left to right: spectra of H2CO 30,3−20,2, HCO 31,3−21,2, HDCO 32,1−22,0, 32,2−22,1, 31,3−21,2, 30,3−20,2 and D2CO 30,3−20,2 observed toward AFGL5142, IRAS 05358+3543, and G5.89-0.39 using the IRAM 30 m telescope (Fontani et al. 2011, 2015) and the APEX SEPIA Band 5 receiver (this work). Red and blue lines indicate the fitted different velocity components for AFGL5142 and G5.89-0.39 with fixed velocity and width. The C13CS N = 15−14 and J = 15−14 transition is shown in the fourth panels from the top.

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.