EDP Sciences logo
Subscriber Authentication Point
Search
Menu
Home All issues Volume 542 (June 2012) A&A, 542 (2012) L15 Full HTML
Issue
A&A
Volume 542, June 2012
GREAT: early science results
Article Number L15
Number of page(s) 4
Section Letters
DOI https://doi.org/10.1051/0004-6361/201218927
Published online 10 May 2012

© ESO, 2012

1. Introduction

The earliest phases of massive star formation (MSF) are still poorly understood. We know that massive stars are born in dense clumps within giant molecular cloud complexes. Ultracompact HII regions (UCHIIRs) embedded within these clumps represent a key phase in the early lives of massive stars (see review by Hoare 2005). In their environments, often hot (>100 K), compact (<0.1 pc), and dense cores are found, some of which are believed to be in a stage prior to the formation of UCHIIRs (Kurtz et al. 2000). More recently, so-called high-mass protostellar objects (HMPOs) or massive young stellar objects (MYSOs) were recognized to likely represent an even earlier stage of MSF (e.g. Beuther et al. 2002). Very recently, clumps within infrared dark clouds (IRDCs) were found to be promising candidates for even earlier stages in the formation of massive stars (see Menten et al. 2005, and references therein), sometimes called pre-protocluster clumps.

Table 1

Ammonia source sample and observing parameters.

One of the important results of studies of low-mass star formation has been the observation of infall motions (e.g. Belloche et al. 2002), which give direct evidence of accretion. Models of high-mass cluster formation also predict a large-scale contraction as the protocluster evolves and molecular gas is funneled from the outer regions of the core into the center of the cluster (Bonnell & Bate 2002; Bonnell et al. 2004). But toward high-mass star-forming cores, the observational evidence of infall is still scarce (e.g. Wu & Evans 2003).

Most studies rely on the observations of so-called “blue-skewed” line profiles: with a negative excitation gradient in a molecular clump, molecular lines with suitable optical depths (e.g. low J rotational lines of HCO+) will show a central self-absorption dip whose blue peak is stronger than the red peak. But there are many pitfalls in the interpretation of this signature (see discussion in Evans 2003). Especially at distances (few kpc) typical of MSF sites, other kinematical properties such as rotation and outflow activity might easily be mistaken for infall. Furthermore, abundance variations in the clumps can change the self-absorption profile tremendously. Still, this method led to evidence for infalling envelopes of quite a few of hot cores (Wyrowski et al. 2006).

Alternatively, absorption lines toward compact and strong continuum sources can be observed. In particular, centimeter ammonia inversion lines have been successfully detected in absorption toward bright UCHIIRs, e.g. by Sollins et al. (2005) in G10.6-0.4. They derive gas velocities consistent with material spiraling toward the central newly-formed stars. Similar results have been reported by Zhang & Ho (1997) in the W51 massive star-forming region (see their Fig. 6), and by Beltrán et al. (2006) toward G24.78+0.08. These results strongly indicate the presence of infall, as predicted by the accretion models. But this method is restricted to a fairly late state of MSF when an UCHII has already developed and can be used as a background source.

At THz frequencies new opportunities arise now: their high dust column densities make MSF clumps very bright far-infrared (FIR) continuum sources, especially in their early stages. Therefore, molecular lines with lower excitation than the (background) continuum temperature will be in absorption and will probe the kinematics of the clump on the line of sight toward the deeply embedded forming massive clusters. This signature is then a much more direct probe of infall and can be used through all embedded stages of MSF.

Of particular interest is the ammonia molecule NH3 , which has been extensively studied, mostly via its centimeter inversion lines, and proven to be one of the best-known interstellar “thermometers” (Ho & Townes 1983; Maret et al. 2009). In the THz range rotational transitions of ammonia can now be observed, in particular transitions from non-metastable to metastable levels, which will have low excitation temperatures and therefore are likely to be observed in absorption. In addition, ammonia is known to be one of the molecules that is the least likely to freeze out in the earliest, coldest stages of molecular clumps (e.g. Bergin & Langer 1997).

We therefore present here a first small survey of sources in a range of evolutionary stages that we observed with the Stratospheric Observatory For Infrared Astronomy (SOFIA) in ammonia to measure their infall signatures. This observational evidence of accretion in the clumps harboring the forming massive (proto) clusters is essential for our understanding of the MSF process.

2. Observations and data reduction

We used the GREAT1 instrument (Heyminck et al. 2012) onboard SOFIA to observe the NH3 32+ − 22− line at 1810.379971 GHz (taken from the JPL line catalog, Pickett et al. 1998). This line has the same lower level as the well-studied centimeter inversion line at 23.7 GHz of the (J,K) = (2,2) level, hence we know that this level is sufficiently populated to give rise to strong absorption signatures.

To reach this frequency, a slight technical change was necessary: the modularity of GREAT allows the implementation of new technical opportunities between flights. A spare local oscillator (LO) source for the high-frequency channel allowed tuning to slightly lower than advertised frequencies (1810 instead of 1820 GHz with the nominal LO). With this setup, the NH3 line became accessible, but only if tuned into the lower sideband of the receiver. GREAT operates in double sideband mode and parts of the image sideband then cover a region of very low transmission, which led to uncertainties in the determination of the (weak) continuum level. The lower frequency channel was tuned to the CO (13–12) line at 1496.922909 GHz and was in turn used, at much better transmission, to determine the THz continuum flux of the sources.

The beam sizes at the observing frequencies are 20 and 16 arcsec, respectively. The three sources given in Table 1 were observed on 2011 July 20 at an altitude ranging from 39 000 to 43 000 feet.

The system temperatures during the observations and integration times are given in Table 1. As backends, MPIfR fast Fourier transform spectrometers (Klein et al. 2012) were used to cover a bandwidth of 1.5 GHz for each channel with a velocity resolution of 0.03 km s-1, which was later lowered to 1 km s-1 to increase the signal-to-noise ratio in the spectra. The wobbling secondary was chopped with a throw of 120′′ about the cross elevation axis in a symmetric mode. The pointing was monitored with the fiber optic gyroscopes. Spectra were calibrated to a scale and then the conversion to TMB was made assuming a forward efficiency of 0.95 and beam efficiencies of 0.54 and 0.51 for the lower and higher frequency receivers, respectively.

In addition, complementary ground-based spectra of the C17O (3–2) and HNC (4–3) were obtained with the APEX telescope.

Final processing and analysis of the data was then performed using the CLASS program within the GILDAS software2.

thumbnail Fig. 1

NH3 32 +  − 22 −  spectra of the observed sources. Results of Gaussian fits to the line are overlaid in green. The systemic velocities of the sources, determined using C17O (3–2), are shown with dotted lines.

3. Results

The averaged and baseline-subtracted spectra are shown in Fig. 1. Ammonia absorption lines were detected in all three sources. The line parameters from Gaussian fits are given in Table 2. Compared with C17O (3–2) line velocities, which as optically thin emission lines probe the systemic velocities of the sources with a comparable resolution of 18′′, all ammonia absorption lines are redshifted. This is consistent with blue-skewed emission profiles observed in HNC (4–3) toward these clumps with the APEX telescope. The fact that we observe absorption lines is not surprising: compared to the microwave inversion lines, the FIR rotational lines have Einstein A values about five magnitudes higher, hence they will decay rapidly from the non-metastable (J > K) into the metastable (J = K) levels, which in turn only decay slowly via collisions (Ho & Townes 1983). Thus, the metastable levels are “overpopulated” compared with the non-metastable ones.

Because we were unable to measure the 1.8 THz continuum reliably owing to the double sideband reception and the low atmospheric transmission in part of the image band, the 1.5 THz continuum was used to estimate the continuum level at the NH3 frequency. Table 3 gives the 1.5 THz continuum level, which was divided by 2 to correct for the single sideband calibration that by default is applied to the lines. The modeling of the clumps described in the next section suggests that the real 1.8 THz continuum might be up to 20% higher. The optical depths of the absorption can then be estimated from the line-to-continuum ratio as τ =  −log (1 + TMB/Tcont) using the temperatures from Tables 2 and 3. Optical depths on the order of unity are found, meaning that the filling factor of the absorbing material in front of the dust continuum must be very high.

Table 2

Line parameters from Gaussian fits to the NH3 lines.

Table 3

Continuum temperatures at 1.5 THz, the NH3 optical depths and the column densities in the (2, 2) levels.

The total column density of the absorbing NH3 can then be computed from (e.g. Comito et al. 2003), (1)This equation is valid for excitation temperatures much lower than the continuum temperature and also much lower than the “transition temperature” Tul (hν/k = 86 K for this NH3 transition), which likely is the case given the rapid radiative decay of the non-metastable upper level. The transition probability of Aul = 3.56 × 10-2 s-1 leads to a critical density for the transition of 109 cm-3. gl and gu are the statistical weights of the lower and upper levels, respectively. Both inversion levels have almost the same population for densities above 104 cm-3, hence the total column density in the (2, 2) levels is twice as high. This leads to column densities of 1−2 × 1014 cm-2 as given in Table 3.

thumbnail Fig. 2

G34.26+0.15 SOFIA NH3 spectrum compared with the VLA NH3 (2, 2) spectrum (shown in red) taken from Gómez et al. (2000), which was integrated over the region that shows absorption. A two-component hyperfine fit to the (2, 2) spectrum is shown in green.

In Fig. 2 we show a comparison of the (2, 2) inversion line seen with the VLA (Keto et al. 1987; Gómez et al. 2000) with the SOFIA results. This inversion line shares the same lower level with the line observed by SOFIA. The velocity and width of the two absorption lines agree very well. Interestingly, the VLA spectrum shows the (2, 2) line’s outer hyperfine structure components in emission. This can be understood when the absorbing continuum is only filling the beam partly and the absorption spectrum is blended with an additional high optical depth emission component from the hot core that is situated offset from the continuum.

4. Analysis

To model the observed absorption lines, we used the physical structures determined by Rolffs et al. (2011): they fitted the continuum structure of MSF clumps as observed in the ATLASGAL survey (Schuller et al. 2009) with density power laws n ~ r-1.5 (for our observed sources) and determined temperature structures dictated by the inner heating sources. The velocity structure was assumed to be a fixed fraction of the free-fall velocities and adjusted using RATRAN (Hogerheijde & van der Tak 2000) to fit high-density molecular probes observed with the APEX telescope. We compared the predicted 1.5 THz continuum of the models with the observed ones. G31.41 was fitted well with dust of Ossenkopf & Henning (1994) using their models with coagulation at n = 105 cm-3 but no ice mantles. For G34.26 we adjusted the grain properties to dust with thin mantles that have a higher spectral index to reproduce the high observed continuum temperature. W43-MM1 was not modeled by Rolffs et al. (2011). For this source we scaled the G31.41 model to match the observed 1.5 THz continuum flux.

In the next step we adjusted the ammonia abundance (n(H2)/n(NH3)) and the velocity to fit the SOFIA NH3 absorption lines with the RATRAN code. While in principle the clumps could have an increased hot core NH3 abundance in the inner T > 100 K region, this hot core component to the spectrum is negligible because of its small filling factor. To fit the velocity, we adjusted the turbulence widths of the gas and the free-fall fraction f in , where Min is the mass inside r (cf. Rolffs et al. 2011). An example fit is shown in Fig. 3 for G34.26+0.15 together with a fit of the model to the NH3 (2, 2) inversion line observed by Churchwell et al. (1990). In the model, the velocity was adjusted to 30% of the free-fall velocity and an ammonia abundance of 0.8 × 10-8. The Doppler b-parameter was adjusted to b = 1.5 km s-1. In the SOFIA spectrum, there is even evidence for additional excess redshifted absorption, not fit by the model, albeit only with low signal-to-noise ratio. The optically thick main cm-inversion line is somewhat overpredicted, whereas the outer hyperfine satellites are reproduced quite well. High infall velocities in the innermost part of the model do not contribute to the observed line profile, since the filling factor of the hot core region is too low. W43-MM1 and G31.41 could be fitted with velocity fields with 20% free-fall contribution.

thumbnail Fig. 3

G34.26+0.15 NH3 SOFIA and Effelsberg (taken from Churchwell et al. 1990) spectra in red compared with the RATRAN model discussed in the text (black profiles). The systemic velocity from C17O is indicated by a dotted line.

5. Discussion and conclusions

From the observed and modeled velocity fields the infall rate of the clumps can be determined. From the RATRAN models, Ṁ was determined from the model density and velocity at the radius corresponding to the SOFIA beam size. Alternatively, Eq. (4) from Beltrán et al. (2006) can be used with the observed column density and infall velocity. Both approaches agree and lead to infall rates between 3−10 × 10-3 M⊙ yr-1. These rates are high enough to quench any HII region (Walmsley 1995) and to overcome the radiation pressure from the forming (proto) cluster, although we only probed the infall at the outer envelope. In the inner part accretion might already be halted. From the absorption, we also only probed the infall on the line-of-sight, hence it might not be spherical.

In summary, we present the first velocity-resolved ammonia spectra from a rotational transition at THz frequency toward

MSF clumps. Absorption of a comparable NH3 32− − 22+ line at 1764 GHz was also recently observed with Herschel in the low-mass source IRAS16293-2422 (Hily-Blant et al. 2010). In G34.26 the redshifted absorption agrees with cm absorption observed with the VLA toward the UCHII component B (Heaton et al. 1989). This component is within the SOFIA beam but the strong absorption we detected implies that the absorption fills most of the dust continuum, which peaks toward the hot molecular core. G31.41 and W43-MM1 do not have compact cm continuum that might be used for absorption studies at cm wavelengths, hence our observations are first detections of redshifted absorption in these sources and demonstrate that ammonia absorption in front of strong dust continuum can be successfully used to probe infall in a variety of evolutionary stages of MSF clumps.

1

GREAT is a development by the MPI für Radioastronomie and the KOSMA/Universität zu Köln, in cooperation with the MPI für Sonnensystemforschung and the DLR Institut für Planetenforschung.

Acknowledgments

Based [in part] on observations made with the NASA/DLR Stratospheric Observatory for Infrared Astronomy. SOFIA Science Mission Operations are conducted jointly by the Universities Space Research Association, Inc., under NASA contract NAS2-97001, and the Deutsches SOFIA Institut under DLR contract 50 OK 0901.

References

  1. Belloche, A., André, P., Despois, D., & Blinder, S. 2002, A&A, 393, 927 [Google Scholar]
  2. Beltrán, M. T., Cesaroni, R., Codella, C., et al. 2006, Nature, 443, 427 [NASA ADS] [CrossRef] [PubMed] [Google Scholar]
  3. Bergin, E. A., & Langer, W. D. 1997, ApJ, 486, 316 [NASA ADS] [CrossRef] [Google Scholar]
  4. Beuther, H., Schilke, P., Menten, K. M., et al. 2002, ApJ, 566, 945 [NASA ADS] [CrossRef] [Google Scholar]
  5. Bonnell, I. A., & Bate, M. R. 2002, MNRAS, 336, 659 [NASA ADS] [CrossRef] [Google Scholar]
  6. Bonnell, I. A., Vine, S. G., & Bate, M. R. 2004, MNRAS, 349, 735 [NASA ADS] [CrossRef] [Google Scholar]
  7. Churchwell, E., Walmsley, C. M., & Cesaroni, R. 1990, A&AS, 83, 119 [NASA ADS] [Google Scholar]
  8. Comito, C., Schilke, P., Gerin, M., et al. 2003, A&A, 402, 635 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  9. Evans, II, N. 2003, in SFChem 2002: Chemistry as a Diagnostic of Star Formation, ed. C. L. Curry, & M. Fich, 157 [Google Scholar]
  10. Gómez, Y., Rodríguez-Rico, C. A., Rodríguez, L. F., & Garay, G. 2000, Rev. Mex. Astron. Astrofis., 36, 161 [Google Scholar]
  11. Heaton, B. D., Little, L. T., & Bishop, I. S. 1989, A&A, 213, 148 [NASA ADS] [Google Scholar]
  12. Heyminck, S., Graf, U., Güsten, R., et al. 2012, A&A, 542, L1 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  13. Hily-Blant, P., Maret, S., Bacmann, A., et al. 2010, A&A, 521, L52 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  14. Ho, P. T. P., & Townes, C. H. 1983, ARA&A, 21, 239 [NASA ADS] [CrossRef] [Google Scholar]
  15. Hoare, M. G. 2005, Ap&SS, 295, 203 [NASA ADS] [CrossRef] [Google Scholar]
  16. Hogerheijde, M. R., & van der Tak, F. F. S. 2000, A&A, 362, 697 [NASA ADS] [Google Scholar]
  17. Keto, E. R., Ho, P. T. P., & Reid, M. J. 1987, ApJ, 323, L117 [NASA ADS] [CrossRef] [Google Scholar]
  18. Klein, B., Hochgürtel, S., Krämer, I., et al. 2012, A&A, 542, L3 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  19. Kurtz, S., Cesaroni, R., Churchwell, E., Hofner, P., & Walmsley, C. M. 2000, Protostars and Planets IV, 299 [Google Scholar]
  20. Maret, S., Faure, A., Scifoni, E., & Wiesenfeld, L. 2009, MNRAS, 399, 425 [NASA ADS] [CrossRef] [Google Scholar]
  21. Menten, K. M., Pillai, T., & Wyrowski, F. 2005, in Massive Star Birth: A Crossroads of Astrophysics, ed. R. Cesaroni, M. Felli, E. Churchwell, & M. Walmsley, IAU Symp., 227, 23 [Google Scholar]
  22. Motte, F., Schilke, P., & Lis, D. C. 2003, ApJ, 582, 277 [NASA ADS] [CrossRef] [Google Scholar]
  23. Ossenkopf, V., & Henning, T. 1994, A&A, 291, 943 [NASA ADS] [Google Scholar]
  24. Pickett, H., Poynter, R. L., Cohen, E. A., et al. 1998, J. Quant. Spec. Radiat. Transf., 60, 883 [Google Scholar]
  25. Rolffs, R., Schilke, P., Wyrowski, F., et al. 2011, A&A, 527, A68 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  26. Schuller, F., Menten, K. M., Contreras, Y., et al. 2009, A&A, 504, 415 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  27. Sollins, P. K., Zhang, Q., Keto, E., & Ho, P. T. P. 2005, ApJ, 624, L49 [NASA ADS] [CrossRef] [Google Scholar]
  28. Walmsley, M. 1995, in Rev. Mex. Astron. Astrofis. Conf. Ser., 1, ed. S. Lizano, & J. M. Torrelles, 137 [Google Scholar]
  29. Wu, J., & Evans, II, N. J. 2003, ApJ, 592, L79 [NASA ADS] [CrossRef] [Google Scholar]
  30. Wyrowski, F., Heyminck, S., Güsten, R., & Menten, K. M. 2006, A&A, 454, L95 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  31. Zhang, Q., & Ho, P. T. P. 1997, ApJ, 488, 241 [NASA ADS] [CrossRef] [Google Scholar]

All Tables

Table 1

Ammonia source sample and observing parameters.

In the text
Table 2

Line parameters from Gaussian fits to the NH3 lines.

In the text
Table 3

Continuum temperatures at 1.5 THz, the NH3 optical depths and the column densities in the (2, 2) levels.

In the text

All Figures

thumbnail Fig. 1

NH3 32 +  − 22 −  spectra of the observed sources. Results of Gaussian fits to the line are overlaid in green. The systemic velocities of the sources, determined using C17O (3–2), are shown with dotted lines.
In the text
thumbnail Fig. 2

G34.26+0.15 SOFIA NH3 spectrum compared with the VLA NH3 (2, 2) spectrum (shown in red) taken from Gómez et al. (2000), which was integrated over the region that shows absorption. A two-component hyperfine fit to the (2, 2) spectrum is shown in green.
In the text
thumbnail Fig. 3

G34.26+0.15 NH3 SOFIA and Effelsberg (taken from Churchwell et al. 1990) spectra in red compared with the RATRAN model discussed in the text (black profiles). The systemic velocity from C17O is indicated by a dotted line.
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.