Herschel observations of the hydroxyl radical (OH)
in young stellar objects††thanks: Herschel is an ESA space observatory with science instruments provided by European-led Principal Investigator consortia and with important participation from NASA.
Key Words.:Astrochemistry — Stars: formation — ISM: molecules — ISM: jets and outflows — ISM: individual objects: HH 46
Aims:‘Water in Star-forming regions with Herschel’ (WISH) is a Herschel Key Program investigating the water chemistry in young stellar objects (YSOs) during protostellar evolution. Hydroxyl (OH) is one of the reactants in the chemical network most closely linked to the formation and destruction of HO. High-temperature () chemistry connects OH and HO through the OH + H HO + H reactions. Formation of HO from OH is efficient in the high-temperature regime found in shocks and the innermost part of protostellar envelopes. Moreover, in the presence of UV photons, OH can be produced from the photo-dissociation of HO through HO + OH + H.
Methods:High-resolution spectroscopy of the 163.12m triplet of OH towards HH 46 and NGC 1333 IRAS 2A was carried out with the Heterodyne Instrument for the Far Infrared (HIFI) on board the Herschel Space Observatory. The low- and intermediate-mass protostars HH 46, TMR 1, IRAS 15398-3359, DK Cha, NGC 7129 FIRS 2, and NGC 1333 IRAS 2A were observed with the Photodetector Array Camera and Spectrometer (PACS) on Herschel in four transitions of OH and two [O i] lines.
Results:The OH transitions at 79, 84, 119, and 163m and [O i] emission at 63 and 145m were detected with PACS towards the class I low-mass YSOs as well as the intermediate-mass and class I Herbig Ae sources. No OH emission was detected from the class 0 YSO NGC 1333 IRAS 2A, though the m was detected in absorption. With HIFI, the 163.12m was not detected from HH 46 and only tentatively detected from NGC 1333 IRAS 2A. The combination of the PACS and HIFI results for HH 46 constrains the line width (FWHM 11km s) and indicates that the OH emission likely originates from shocked gas. This scenario is supported by trends of the OH flux increasing with the [O i] flux and the bolometric luminosity, as found in our sample. Similar OH line ratios for most sources suggest that OH has comparable excitation temperatures despite the different physical properties of the sources.
The hydroxyl radical (OH) is a cornerstone species of the oxygen chemistry in dense clouds and is particularly important in the chemical reaction network of water. HO and OH are closely linked through the OH + H HO + H reactions. The formation path of HO from OH is efficient at the high temperatures found in shocks or in the innermost parts of circumstellar envelopes (Kaufman & Neufeld, 1996; Charnley, 1997). Below about 250 K, standard gas-phase chemistry applies, in which HO is formed and destroyed through ion-molecule reactions. In regions not completely shielded from UV radiation, photo-dissociation becomes a major destruction path of HO, leaving OH as a byproduct. A better understanding of the OH emission will therefore help to constrain the water chemistry.
The observation of far-infrared (FIR) rotational OH lines by ground based facilities is severely limited by the Earth’s atmosphere. Previous studies of OH FIR emission with the Infrared Space Observatory (ISO) showed that OH is one of the major molecular coolants in star-forming regions (e.g. Giannini et al., 2001). However, with a large beam of 80, ISO was unable to resolve the central source from the outflows, preventing an assessment of the origin of the OH emission. Interpretation of the ISO OH measurements thus relied mostly on the assumption that the OH emission arises from gas with the same temperature and density as the high-J CO FIR emission (e.g. Nisini et al., 1999; Ceccarelli et al., 1998). The Herschel Space Observatory permits observations of OH FIR transitions at both higher angular and spectral resolution and at higher sensitivity than ISO.
Observations of HO, OH and related species towards a large set of young stellar objects over a wide range of luminosities and masses are being carried out in the ‘Water In Star-forming regions with Herschel’ (WISH) key program to trace the water chemistry during protostellar evolution (van Dishoeck et al. in prep). OH emission at 163.12m (1837.8 GHz) was detected with PACS towards the class I YSO HH 46 (van Kempen et al., 2010b), but the triplet, split by 90 MHz, could not be resolved. Based on modeling results, the OH emission was attributed to a -type shock and not to the quiescent envelope. We carried out high-resolution spectroscopy with HIFI to test whether the detected OH emission is dominated by shock contribution by resolving the line profiles. This paper presents the HIFI observations in HH 46 as well as the class 0 object NGC 1333 IRAS 2A. New PACS observations of OH and [O i] are reported for the low-mass protostars IRAS 15398-3359, NGC 1333 IRAS 2A, and TMR 1.
2 Observations and data reduction
Because of spin-orbit interaction, the OH rotational levels are built within two ladders, and (Offer & van Dishoeck, 1992). Each level is further split by doubling and hyperfine structure. A level diagram can be found in Fig. 6 in the appendix. We use the molecular data from the Leiden atomic and molecular database LAMDA 111http://www.strw.leidenuniv.nl/moldata/ (Schöier et al., 2005).
High-resolution observations of the OH triplet at 163.12m (1837.747, 1837.817 and 1837.837 GHz) were performed with the Heterodyne Instrument for the Far-Infrared (HIFI, De Graauw et al., 2010) on board the ESA Herschel Space Observatory (Pilbratt et al. 2010) towards HH 46 and NGC 1333 IRAS 2A. HIFI data were stitched together using the Herschel Interactive Processing Environment (HIPE v3.0.1, Ott, 2010) and further analyzed using GILDAS-CLASS222http://www.iram.fr/IRAMFR/GILDAS software. We removed standing waves after the subtraction of a low-order polynomial and calibrated to scale using a main beam efficiency of 0.74. The H and V polarizations were combined.
HH 46, TMR 1, IRAS 15398-3359, and NGC 7129 FIRS 2 were observed with PACS (Poglitsch et al. 2010) in line spectroscopy mode around four OH doublets at 79, 84, 119, and m. The [O i] 63 and 145.5m lines were also observed except for TMR 1. Each segment at m and m covers 1 and m at and , respectively. In addition, DK Cha and NGC 1333 IRAS 2A were observed with PACS from 55–210 m at in range spectroscopy mode. Details of the PACS observations of HH 46, NGC 7129 FIRS 2, and DK Cha are described in van Kempen et al. (2010b), Fich et al. (2010), and van Kempen et al. (2010a), respectively. All spectra were reduced with HIPE v2.9. PACS spectra are recorded in a array of square spatial pixels (spaxels). The observations of IRAS 15398-3359, TMR 1, NGC1333 IRAS 2A, and DK Cha were mispointed sufficiently in a way that the peak of the continuum emission differs from the central spaxel.
The line and continuum emission seen by PACS is spatially extended in most cases. Spectra were extracted from the spaxels that include OH emission. The wavelength-dependent point-spread function was roughly corrected by comparing the amount of continuum emission in the summed spaxels to that in the total array. Most integrated line intensities were derived from Gaussian fits to the unresolved lines. In a few cases, the OH doublets could not be resolved and the intensities were then derived by simple integration over the spectrum. The absolute flux calibration below and above m was separately determined from in-flight observations of (point) calibration sources. The relative spectral response function within each band was determined from ground calibration prior to launch. The uncertainty in absolute and relative fluxes is estimated to be 30-50%. Additional details on the observations can be found in the appendix.
HIFI did not detect the 163.12m OH hyperfine triplet at the noise level of mK on scale for the nominal WBS resolution of about 1.1 MHz (). The spectrum is presented in Fig. 1. Combining the HIFI and PACS observations constrains the line width (see Sect. 4).
PACS detected OH emission at , and m towards the class I sources HH 46, TMR 1 and IRAS 15398-3359 as well as the class I Herbig Ae star DK Cha and the intermediate-mass source NGC 7129 FIRS 2 (Fig. 2). An exception is the m line of HH 46, where we only have an upper limit because of the uncertain baseline towards the end of the spectral window where the line is located. The [O i] m and m lines were also detected. An overview of the results is given in Table 1. The 1 errors listed in Table 1 do not include the systematic error of 30-50% from uncertainty in the calibration. Because of blending with CO(31-30), the m doublet component of OH is not listed. For the analysis, we assumed the flux to be identical to the m OH flux because all observed doublets show comparable fluxes of the components within the calibration uncertainty.
The only class 0 YSO in the sample, NGC 1333 IRAS 2A, is fundamentally different from all other sources in the sample. The OH m doublet is seen in absorption, with an equivalent width of about for each component. No other OH lines were detected at the noise level obtained after removal of the fringing effects. However, the upper limits are larger than the fluxes detected from the class I sources. On the other hand, the upper limit on the [O i] m emission is at least a factor of four lower than the weakest [O i] m line found in our sample.
To constrain the line width of the m OH triplet from HH 46, the HIFI non-detection is combined with the flux () of the unresolved lines measured with the high-sensitivity PACS instrument. The expected integrated intensity for HIFI is for the triplet in total, as calculated from the PACS observation. This value was derived for a beam under the assumption that the source is unresolved, because the OH emission from HH 46 appears to be centrally concentrated (van Kempen et al., 2010b). Assuming LTE ratios () for the three components, the strongest transition has an integrated intensity of . An upper limit on the integrated intensity is calculated from the HIFI observation using with the velocity resolution, the expected line width and an assumed calibration uncertainty. Rebinning the spectrum to a resolution of 16 MHz () yields an rms of mK and would therefore allow () detections for Gaussian lines with (). From the non-detection we conclude that the flux observed with PACS is likely to be emitted from lines with FWHM . For comparison, the FWHM of the HO transition derived from recent unpublished HIFI observations is about . The CO(6-5) and CO(7-6) lines observed with APEX by van Kempen et al. (2009) are narrower.
|HH 46||TMR 1||IRAS 15398||DK Cha||NGC 7129||N 1333 I 2A|
|OH||79.12||3789.3||181.9||55 7||128 14||194 29||360 87||229 33||350|
|OH||79.18||3786.3||181.7||38 5||102 14||-||315 58||230 39||-|
|OH||84.60||3543.8||290.5||87 6||170 18||93 16||347 48||181 28||165|
|OH||119.23||2514.3||120.7||38 9||83 10||121 17||95 50||132 39||absorption|
|OH||119.44||2510.0||120.5||44 7||101 11||130 17||-||134 59||absorption|
|OH||163.12||1837.8||270.1||22 4||56 8||47 9||170 60||116 45||100|
|OH||163.40||1834.7||269.8||27||55 7||54 10||-||69 26||-|
|O i P–P||63.18||4744.8||227.7||1260 54||-||1958 197||3128 333||1219 130||320|
|O i P–P||145.53||2060.1||326.6||82 8||-||122 17||170 72||104 35||180|
Doublet not resolved. Table lists the integrated flux over both components.
In a similar HIFI observation, OH emission was tentatively detected below the 5 level for the strongest triplet component in NGC 1333 IRAS 2A, in agreement with the upper limits on OH emission in the PACS observation of the same source. Given the uncertain baseline, this needs to be treated with caution.
Figure 3 compares the OH fluxes from YSOs as measured here with PACS and in ISO observations of two additional sources, Ser SMM1 (Larsson et al., 2002) and NGC 1333 IRAS 4 (Giannini et al., 2001). Despite the wide range of luminosities and masses covered in our sample, the sources show surprisingly similar characteristics in terms of their OH emission: the OH m doublet is generally the strongest of the four, while the integrated intensity of the weakest doublet at m is roughly a factor of three lower. The m flux can vary compared to higher excitation lines because this line, which is connected to the ground state level, becomes more easily optically thick. It can also be affected by absorption against the continuum by cold gas layers lying in front of the source. The m doublet links the lowest energy states of both rotational ladders, but because of the smaller Einstein coefficients of the cross-ladder transitions, these lines are less affected by optical depth than the m lines. The m fluxes are lower than those at m for most sources, but higher than at m.
Flux ratios of the observed OH lines were calculated from the summed doublet components to compare the excitation of the involved energy levels among the sources. The line ratios of different sources agree within a factor of two except when the 119m transition is involved (Fig. 7 in the appendix) and are consistent with the results from ISO observations from the class 0 sources NGC 1333 IRAS 4 and Ser SMM1.
Emission in the m transition, which has , indicates that OH is tracing the warm () and dense () gas in our sources. Modeling results show that transitions in the ladder are mostly excited by collisions while the population of the levels is dominated by radiative pumping via the cross-ladder transitions. The weak m lines and emission at m indicate that FIR pumping is less important than collisional excitation in our sources.
Figure 4 shows the trend of stronger OH emission with increasing bolometric source luminosity found in our source sample. We calculated the OH line luminosity individually for each transition and source from the observed fluxes and found indications that the differences between the sources may depend on the individual . The correlation between and is reminiscent of that found for the CO outflow force with luminosity (Bontemps et al., 1996). The latter relation was taken as evidence that more massive envelopes have higher accretion rates and thus drive more powerful outflows. van Kempen et al. (2010b) speculate that the OH emission originates in the wake of the jet impinging on the dense, inner parts of the envelope, creating dissociative shocks in which [O i] is the dominant coolant, followed by OH (Neufeld & Dalgarno, 1989). The relation between OH emission and luminosity supports this scenario.
Comparison of OH with [O i] emission shows that stronger OH emission coincides with higher [O i] intensities (Fig. 5) and also with increasing [O i] 63/145m line flux ratios. For HH 46, TMR 1, and NGC 7129 FIRS 2, the bulk of the [O i] and OH emission comes from close to the protostar where densities are on the order of or higher, as illustrated by the lack of extended OH emission in HH 46 (van Kempen et al., 2010b). Some spatially extended OH emission is detected from IRAS 15398-3359 and DK Cha, and is highly correlated with the spatial extent of [O i] emission. The correlation between the intensities (Fig. 5) suggests that the bulk of [O i] and OH emission originates in the same physical component in all sources. This argument, together with the OH – and OH – [O i] relations, supports the dissociative shock scenario. The [O i] 63m / 145m line ratios are in the range of 13–19, also consistent with fast, dissociative -type shocks ( 60 km s, Neufeld & Dalgarno, 1989). Note that an OH – [O i] relation can also be indicative of photo-dissociation, as argued for Sgr B2 by Goicoechea et al. (2004). Models of photon-dominated regions (Kaufman et al., 1999) predict similar [O i] 63m / 145m line ratios, but those require , which is inconsistent with an origin in the inner, dense envelope.
The OH hyperfine transition triplet at 163.12m (1837.8 GHz) was not detected above the noise level obtained with HIFI. Combined with the flux derived from the unresolved line observed with PACS, this constrains the line width to FWHM . This width is much broader than expected for the quiescent envelope from ground-based observations (van Kempen et al., 2009) and indicates that the observed OH emission most likely stems from shocked gas in HH 46.
Herschel PACS observations of OH lines at , and m have been carried out for four low-mass YSOs, one intermediate-mass protostar and one class I Herbig Ae object. OH emission is detected in all sources except the class 0 YSO NGC 1333 IRAS 2A, where the OH 119m transitions are observed in absorption and only upper limits can be derived for the other lines. Sources with detected OH emission show surprisingly similar OH line ratios despite the large ranges of physical properties covered in this study, suggesting that OH emission might arise from gas at similar conditions in all sources. Furthermore, we find trends of correlations between OH integrated intensities and [O i] emission as well as bolometric luminosity, consistent with an origin in the wake of dissociative shocks. Given the low number of sources in the sample, confirmation from additional observations is needed.
Further OH observations and modeling should eventually allow the determination of the OH/HO abundance ratio in shocks, which traces the UV field through its dependence on the fraction of atomic to molecular hydrogen.
- Bontemps et al. (1996) Bontemps, S., Andre, P., Terebey, S., & Cabrit, S. 1996, A&A, 311, 858
- Ceccarelli et al. (1998) Ceccarelli, C., Caux, E., White, G. J., et al. 1998, A&A, 331, 372
- Charnley (1997) Charnley, S. B. 1997, ApJ, 481, 396
- De Graauw et al. (2010) De Graauw, T., Helmich, F. P., Philipps, T. G., et al. 2010, submitted to this issue
- Fich et al. (2010) Fich, M., Johnstone, D., van Kempen, T. A., et al. 2010, A&A in press
- Giannini et al. (2001) Giannini, T., Nisini, B., & Lorenzetti, D. 2001, ApJ, 555, 40
- Goicoechea et al. (2004) Goicoechea, J. R., Rodríguez-Fernández, N. J., & Cernicharo, J. 2004, ApJ, 600, 214
- Kaufman & Neufeld (1996) Kaufman, M. J. & Neufeld, D. A. 1996, ApJ, 456, 611
- Kaufman et al. (1999) Kaufman, M. J., Wolfire, M. G., Hollenbach, D. J., & Luhman, M. L. 1999, ApJ, 527, 795
- Larsson et al. (2002) Larsson, B., Liseau, R., & Men’shchikov, A. B. 2002, A&A, 386, 1055
- Neufeld & Dalgarno (1989) Neufeld, D. A. & Dalgarno, A. 1989, ApJ, 344, 251
- Nisini et al. (1999) Nisini, B., Benedettini, M., Giannini, T., et al. 1999, A&A, 350, 529
- Offer & van Dishoeck (1992) Offer, A. R. & van Dishoeck, E. F. 1992, MNRAS, 257, 377
- Ott (2010) Ott, S. 2010, in ASP Conference Series, Astronomical Data Analysis Software and Systems XIX, Y. Mizumoto, K.-I. Morita, and M. Ohishi, eds., in press.
- Schöier et al. (2005) Schöier, F. L., van der Tak, F. F. S., van Dishoeck, E. F., & Black, J. H. 2005, A&A, 432, 369
- van Kempen et al. (2010a) van Kempen, T. A., Green, J. D., Evans, N. J., et al. 2010a, A&A in press
- van Kempen et al. (2010b) van Kempen, T. A., Kristensen, L. E., Herczeg, G. J., et al. 2010b, A&A in press
- van Kempen et al. (2009) van Kempen, T. A., van Dishoeck, E. F., Güsten, R., et al. 2009, A&A, 501, 633
Acknowledgements.The work on star formation at ETH Zurich is partially funded by the Swiss National Science Foundation (grant nr. 200020-113556). This program is made possible thanks to the Swiss HIFI guaranteed time program. HIFI has been designed and built by a consortium of institutes and university departments from across Europe, Canada and the United States under the leadership of SRON Netherlands Institute for Space Research, Groningen, The Netherlands and with major contributions from Germany, France and the US. Consortium members are: Canada: CSA, U.Waterloo; France: CESR, LAB, LERMA, IRAM; Germany: KOSMA, MPIfR, MPS; Ireland, NUI Maynooth; Italy: ASI, IFSI-INAF, Osservatorio Astrofisico di Arcetri- INAF; Netherlands: SRON, TUD; Poland: CAMK, CBK; Spain: Observatorio AstronÃ³mico Nacional (IGN), Centro de AstrobiologÃa (CSIC-INTA). Sweden: Chalmers University of Technology - MC2, RSS & GARD; Onsala Space Observatory; Swedish National Space Board, Stockholm University - Stockholm Observatory; Switzerland: ETH Zurich, FHNW; USA: Caltech, JPL, NHSC.
Appendix A Observational details
|Source||Distance||Luminosity||Envelope Mass||Type||RA||Dec||Obs. Date||Obs.id|
|[pc]||[L]||[M]||[h m s]|||
|HH 46||450||16||5.1||Class I||08:25:43.9||51:00:36||2009-10-26||1342186315 (PACS)|
|TMR 1||140||3.7||0.12||Class I||04:39:13.7||+25:53:21||2010-03-29||1342192984 (PACS)|
|IRAS 15398-3359||130||0.92||0.5||Class I||15:43:01.3||34:09:15||2010-02-27||1342191302 (PACS)|
|DK Cha||178||29.4||0.03||Herbig Ae||12:53:17.2||77:07:10.6||2009-12-10||1342188039 (PACS)|
|NGC 7129 FIRS 2||1260||500||50||Intermed.-Mass||21:43:01.7||+66:03:23.6||2009-10-26||1342186321 (PACS)|
|NGC 1333 I 2A||235||20||1.0||Class 0||03:28:55.6||+31:14:37||2010-02-13||1342190686 (PACS)|
|Serpens SMM1||415||82.9||8.7||Class 0||-||ISO|
|NGC 1333 I 4||235||5.8||4.5||Class 0||-||ISO|
Heathcote, S., Morse, J. A., Hartigan, P., et al. 1996, AJ, 112, 1141
van Kempen, T. A., van Dishoeck, E. F., Güsten, R. et al. 2009, A&A, 501, 633
Motte, F., Andre, P., & Neri, R. 1998, A&A, 336, 150
Ohashi, N., Hayashi, M., Kawabe, R. et al. 1996, ApJ, 466, 317
Jørgensen, J. K., Schöier, F.L., & van Dishoeck, E.F. 2002, A&A, 389, 908
Murphy, D. C., Cohen, R., & May, J. 1986, A&A, 167, 234
Froebrich, D. 2005, ApJS, 156, 169
van Kempen, T. A., van Dishoeck, E. F., Hogerheijde, M. R. et al., 2009, A&A, 508, 259
Whittet, D. C. B., Prusti, T., Franco, G. A. P., et al. 1997, A&A, 327, 1194
van Kempen, T. A., Green, J. D., Evans, N. J. et al. A&A, 2010, accepted
Shevchenko, V.S. and Yakubov, S.D. 1989, SvA, 33, 370
Johnstone, D., Fich, M., MCoey, C. et al. 2010, this volume
Crimier, N., Ceccarelli, C., Alonso-Albi, T. et al. 2010, arxiv 1005.0947
Hirota, T., Bushimata, T., Choi, Y. K., et al. 2008, PASJ, 60, 37
Jørgensen, J. K., van Dishoeck, E. F., Visser, R., et al. 2009, A&A, 507, 861
Dzib, S., Loinard, L., Mioduszewski, A.J., et al. 2010, ApJ, in press, arXiv1003.5900
Hogerheijde, M. R., van Dishoeck, E. F., Salverda, J. M. et al., ApJ, 513, 350 ( scaled to a distance of 415 pc)
Larsson, B., Liseau, R. & Men’shchikov, A.B. 2002, A&A, 386, 1055
Giannini, T., Nisini, B. & Lorenzetti, D. 2001, ApJ, 555, 40
Table 2 lists the coordinates, observing dates, and the observation identity numbers of our sources along with the assumed distance, the bolometric luminosity, the envelope mass and the source type. For comparison, Ser SMM1 and NGC 1333 IRAS 4 observed with ISO are included as well. For Ser SMM1, we use the average values of the fluxes presented by Larsson et al. (2002). The data for NGC 1333 IRAS 4 are taken from Giannini et al. (2001). Note that for NGC 1333 IRAS 4, we use the luminosity and mass of NGC 1333 IRAS 4A.