EDP Sciences logo
Subscriber Authentication Point
Search
Menu
Home All issues Volume 569 (September 2014) A&A, 569 (2014) L8 Full HTML
Free Access
Issue
A&A
Volume 569, September 2014
Article Number L8
Number of page(s) 4
Section Letters
DOI https://doi.org/10.1051/0004-6361/201424748
Published online 02 October 2014

© ESO, 2014

1. Introduction

Protostellar jets and outflows are a direct consequence of the accretion mechanism in young stellar objects during their earliest phase (e.g. Ray et al. 2007). The interaction between the ejecta and the circumstellar medium occurs via radiative shocks (e.g. Kaufman & Neufeld 1996; Flower & Pineau des Forêts 2010), whose energy is radiated away through emission lines of atomic, ionic, and molecular species. Hot gas at temperatures above 2000 K cools principally through H2 ro-vibrational lines in the near-IR and abundant atomic and ionic species (e.g. Eislöffel et al. 2000; Giannini et al. 2004). Warm gas components at hundreds of Kelvin cool via mid- and far-IR molecular lines, particularly rotational transitions of H2 (at λ ≤ 28 μm) and lines of other molecular species, such as CO and H2O.

Table 1

H2 transitions observed.

Water has a key role in protostellar environments (van Dishoeck et al. 2011). Its abundance with respect to H2 is expected to increase from <10-7 in cold regions up to 3 × 10-4 in warm gas due to the combined effects of sputtering of grain mantles and high-temperature reactions (Hollenbach & McKee 1989; Kaufman & Neufeld 1996; Flower & Pineau des Forêts 2010; Suutarinen et al. 2014). The Herschel Space Observatory revealed the complexity of H2O line profiles (e.g. Codella et al. 2010; Kristensen et al. 2012; Santangelo et al. 2012; Vasta et al. 2012) and showed that H2O emission probes warm (≳300 K) and dense (nH2> 105 cm-3) gas with spatial distribution that resembles the H2 emission (e.g. Nisini et al. 2010; Tafalla et al. 2013; Santangelo et al. 2013). Low H2O abundances are derived in outflows for warm shocked gas, ranging from a few ×10-6 to a few ×10-5 (e.g. Bjerkeli et al. 2012; Santangelo et al. 2013; Tafalla et al. 2013; Busquet et al. 2014). These abundance values are at least an order of magnitude lower than what is expected in warm shocked gas from shock model predictions (e.g. Kaufman & Neufeld 1996; Flower & Pineau des Forêts 2010). Their determinations rely on the assumption that H2O traces the same gas as the spectrally unresolved low-J H2 0–0 lines. Spectrally resolved observations of H2 are thus needed to directly compare the line profiles and finally test this hypothesis.

The Herbig-Haro object HH 54 is located in the nearby star-forming region Chamaeleon II (D = 180 pc, Whittet et al. 1997). The object shows a clumpy appearance, consisting of several arcsecond-scale bright knots. Knee (1992) associates HH 54 with a monopolar blue-shifted CO outflow, whose driving source remains unclear (e.g. Caratti o Garatti et al. 2006; Ybarra & Lada 2009; Bjerkeli et al. 2011). Mid-IR cooling is dominated by pure rotational H2 lines (Cabrit et al. 1999; Giannini et al. 2006; Neufeld et al. 1998, 2006) probing warm gas with a mixture of temperatures in the range 400−1200 K. HH 54 was also observed in several lines of CO and H2O from space and the ground (Liseau et al. 1996; Nisini et al. 1996; Bjerkeli et al. 2009, 2011).

In this letter we present new ESO VLT high-resolution spectroscopic observations of H2 towards HH 54. The observations are complemented with Herschel/HIFI observations of H2O (212 − 101). This unique dataset is used to spectrally resolve the different excitation components observed in H2. We are finally able to identify the H2 counterpart associated with H2O and derive the H2O abundance in the shocked gas directly.

2. Observations

thumbnail Fig. 1

Upper: H2 1−0 S(1) image of HH 54 from NTT/SofI observations (Giannini et al. 2006). The positioning of the slits adopted for VLT/VISIR and CRIRES observations is shown in blue and red. The beam sizes of Herschel H2O (212−101) (magenta circle), CO (15−14) (green, Bjerkeli et al. 2014), and SEST CO (2−1) (black, Bjerkeli et al. 2009, 2011) are displayed. Offsets are relative to: αJ2000 = 12h55m503, δJ2000 = − 76°56′23′′ (Bjerkeli et al. 2011). Lower: Spitzer/IRS H2 0−0 S(4) image of HH 54 (Neufeld et al. 2006). Symbols are the same as in the upper panel.
Open with DEXTER

Our dataset consists of data collected towards HH 54 with ESO facilities (Table 1) and with Herschel. Figure 1 shows the VLT slit positions and Herschel and SEST beam sizes for the observations presented in this paper in comparison with the H2 1−0 S(1) and 0−0 S(4) maps of the region (Giannini et al. 2006; Neufeld et al. 2006).

2.1. VISIR high-resolution mid-IR spectroscopy

On April 2012 we performed spectrally-resolved observations of H2 0−0 S(4) (see Table 1) with VLT/VISIR (Lagage et al. 2004). The slit was positioned on the basis of the Spitzer image (see Fig. 1); it was oriented in a way to encompass knot B, which was covered by the Herschel single-pointing observations of H2O (see Sect. 2.3), and the C1/C2 knots, which correspond to the brightest knot in the Spitzer H2 emission. We conducted our observations by chopping and nodding the telescope off-source, with equal time on both positions. Data reduction and calibration were performed by using the VISIR pipeline recipes (version 3.5.1)1, which provide standard procedures for flat-fielding and background subtraction. A model for the sky emission lines is used by the pipeline for the wavelength calibration. To fit the dispersion relation we employed a second degree polynomial, which provides higher correlation coefficient with respect to the default pipeline linear solution. The uncertainty on the peak velocity is about 3 km s-1, comparable with the spectral pixel. The IRAF package was used for spectra extraction. Only C1/C2 knots are clearly detected with VISIR; knot M is only tentatively detected in the spectral image, whereas knot B is not detected.

2.2. CRIRES high-resolution near-IR spectroscopy

We carried out high-dispersion spectroscopic observations of the H2 0−0 S(9) and H2 1−0 S(1) transitions (Table 1) towards HH 54 with VLT/CRIRES (Kaeufl et al. 2004). Observations were performed between January and February 2014 during director discretionary time. Since only the bright C1/C2 knots were detected by VISIR, the CRIRES slit was oriented in order to cover them (Fig. 1). Chopping and nodding were performed along the slit to minimise the integration time. Data reduction and wavelength calibration were performed with the CRIRES pipeline recipes (version 2.3.1). The wavelength calibration, based on the comparison with a sky emission model, was satisfactory (high correlation coefficient) for the 0−0 S(9). OH emission lines were used to refine the wavelength scale for the 1−0 S(1). The uncertainty associated with peak velocities is ~2.5 km s-1. The IRAF package was used for spectra extraction.

2.3. Herschel/HIFI observations

Single-pointing observations of H2O (212−101) at 1669.9 GHz were performed with the Heterodyne Instrument for the Far Infrared (HIFI, de Graauw et al. 2010) on board Herschel towards HH 54B (see Fig. 1). The reference coordinates are αJ2000 = 12h55m503, δJ2000 = − 76°56′23′′. The observations were carried out in September 20122. The diffraction-limited beam size is ~13′′. The data were processed with the ESA-supported package hipe version 12.0 for calibration. The HebCorrection and fitHifiFringe tasks within hipe were successfully used to remove the electronic standing waves in Band 6, which affected the line. Further data reduction and analysis were performed using the GILDAS3 software. The antenna temperature scale, , was converted into the main-beam temperature scale, Tmb, using main-beam efficiency factor of 0.71 (Roelfsema et al. 2012). The flux calibration uncertainty is around 10%, based on cross-calibration with Herschel/PACS (Bjerkeli et al. 2011). At the velocity resolution of 1 km s-1, the rms noise is 80 mK (Tmb scale).

3. Two velocity components in H2 observations

thumbnail Fig. 2

Upper: H2 0−0 S(4) towards HH 54C1/C2 is compared with 1−0 S(1) and 0−0 S(9). Spectra are normalized to their peak values. The vertical dashed line marks the systemic velocity (vLSR = + 2.4 km s-1, Bjerkeli et al. 2011). The two vertical dotted lines indicate the velocity of: the 0−0 S(4) peak at − 7 km s-1; and the 1−0 S(1) and 0−0 S(9) peaks at −15 km s-1. Lower: H2O (212 − 101), CO (15−14), and CO (2−1) towards HH 54B are compared with H2 0−0 S(4) at HH 54C1/C2. H2O, CO, and H2 are normalized to the peak of the bump feature in CO (2−1).
Open with DEXTER

Velocity centroids of the CRIRES spectra at C1 and C2 knots are consistent within one spectral pixel (<3 km s-1). The two spectra have thus been averaged to compare with VISIR H2 0−0 S(4) extracted at knot C1/C2. The comparison is presented in the upper panel of Fig. 2. A peak velocity of −7 km s-1 is associated with the 0−0 S(4) line, whereas the higher excitation 1−0 S(1) and 0−0 S(9) lines peak at the higher blue-shifted velocity of −15 km s-1. Our spectrally-resolved H2 observations clearly show for the first time that mid-IR and near-IR H2 lines are well separated in velocity, thus representing two distinct velocity components. This suggests that two separate shock components with different excitation conditions are associated with gas peaking at different velocities.

The comparison between H2 0−0 S(4) at HH 54C1/C2 and H2O (212−101), CO (15−14) (Bjerkeli et al. 2014), and CO (2−1) (Bjerkeli et al. 2009, 2011) observations at HH 54B is presented in the lower panel of Fig. 2. The low-J CO lines, in particular CO (2−1), present a two-components profile: a triangular-shaped low-velocity (LV CO, hereafter) component, which peaks at the systemic velocity of the cloud (+2.4 km s-1); and an additional superposed “bump-like” component (Bjerkeli et al. 2011) centred at the blue-shifted velocity of − 7 km s-1. This latter feature seems to dominate the emission of the high-J CO (15−14). The similarity between CO (15−14) and H2O line profiles, taken with similar beam sizes (12′′ and 13′′), suggests that the bump feature is associated with the H2O emitting gas and has higher excitation with respect to the LV gas.

Although the H2 0−0 S(4) spectrum is taken at the C1/C2 knot, the comparison with H2O and high-J CO observed at knot B shows that the three lines trace emission in the same velocity range. Moreover, taking the different spectral resolutions and the uncertainty on the H2 peak velocity determination into account (see Sect. 2.1), the H2 0−0 S(4) line profile well resembles the H2O and high-J CO line profiles (Fig. 2 bottom). HIFI maps of CO (10−9) and lower-J CO lines by Bjerkeli et al. (2011) show that, although the relative intensity of the LV and bump components changes within the HH 54 region, their peak velocities remain constant within 2 km s-1 among the different knots. We thus also assume that the peak velocity of the H2O emission, which appears to be associated with the high-J CO emission at knot B, does not change within the region and in particular along the VISIR slit. In this case, the H2 0−0 S(4) emission would be associated with the same gas as traced by H2O and high-J CO.

In conclusion, our observations detect for the first time the presence of a stratification in velocity in the H2 gas from low- to high-excitation emission lines. The H2 0−0 S(4) component appears to be associated with H2O and high-J CO, as expected from the comparison between the spatial distributions (e.g. Nisini et al. 2010; Tafalla et al. 2013; Santangelo et al. 2013). We note that in the low-J CO an additional gas component around the systemic velocity is detected. This gas component is not observed in the high-J CO lines and in the H2 lines, since higher temperatures are needed to excite them. On the other hand, the higher velocity component associated with H2 1−0 S(1) and 0−0 S(9) is not detected in the CO emission, even in the higher-J lines, since it is associated with a gas at even higher temperatures (T ≳ 2000 K).

4. H2O abundance estimate

Our new observations allow us to spectrally identify the 0−0 S(4) line as the H2 counterpart associated with H2O, with the assumption that the H2 and H2O profiles do not change between the C and B knots. This can be used to accurately constrain the temperature of the gas from the H2 emission and derive correct H2O abundances with respect to H2. Neufeld et al. (2006) mapped H2 S(0)−S(7) pure rotational lines towards HH 54 with Spitzer/IRS. Their H2 rotational diagram, constructed over a 15′′ region encompassing HH 54B, indicates the presence of warm gas with temperatures in the range 400−1200 K. According to these authors, a temperature range of 700−1000 K is associated with the 0−0 S(4) emission, which corresponds N(H2) = 6.6 × 1019 and 2.1 × 1019 cm-2 over 13′′ for 700 and 1000 K, respectively.

We assumed for the H2O emission the same temperature range as derived from the H2 0−0 S(4) line and a gas density n(H2) ≳ 105 cm-3 (e.g. Tafalla et al. 2013; Santangelo et al. 2013; Busquet et al. 2014). We used the radex molecular LVG radiative transfer code (van der Tak et al. 2007) to model the observed H2O (212 −101) emission. A typical line width of 10 km s-1 was adopted from a Gaussian fitting to the spectrum (see Fig. 2). The lower limit on the H2 density corresponds to an upper limit on the derived column density. In particular, we obtain N(H2O) <3 × 1014 cm-2 over a 13′′ area. The comparison with the H2 column density obtained from the 0−0 S(4) for the same temperature range gives an H2O abundance X(H2O) < 1.4 × 10-5. A lower H2 density of 2 × 104 cm-3 (Bjerkeli et al. 2011, 2014) would increase the H2O abundance by a factor of 2. The upper level energy of H2O (212−101), which is about 114 K, is much smaller than that of the H2 S(4) line (Table 1). Therefore, we cannot exclude that H2O emission is associated with a colder gas component that is not probed by our H2 observations. However, a temperature lower than the assumed 700−1000 K would indicate an even lower H2O abundance, thus strengthening our result. The derived upper limit on the H2O abundance is in agreement with the abundance value of 10-5 derived by Liseau et al. (1996) and Bjerkeli et al. (2011) from ISO and Herschel observations of transitions at similar wavelengths as well as with the upper limit of < 1.6 × 10-4 obtained by Neufeld et al. (2006) based on non-detections of shorter wavelength transitions covered by Spitzer. Our H2O abundance estimate in HH 54 confirms the values recently found by Herschel in outflows from Class 0 sources (e.g. Bjerkeli et al. 2012; Santangelo et al. 2013; Tafalla et al. 2013; Busquet et al. 2014), which are based on the assumption that H2O traces the same gas as traced by the low-J H2 emission.

An estimate of the H2O abundance at the C knot can also be derived using the PACS map of H2O (212 − 101) by Bjerkeli et al. (2011). The H2O flux density at knot C is a factor of 2 lower than at the position of the HIFI H2O observations, which yields N(H2O) < 1014 cm-2. The H2 column density obtained from the 0−0 S(4) at knot C is in the range N(H2) = 5 × 1019−1.6 × 1020 cm-2 for 1000 and 700 K, respectively. The comparison between H2O and H2 indicates an H2O abundance X(H2O) < 2 × 10-6, which is even more strict than that derived at knot B. This indicates a variation of H2O abundance within the HH 54 region, with a decrease towards the peak of the H2 S(4) emission. This may explain the different emission peaks of the H2O distribution observed by PACS (Bjerkeli et al. 2011) and the H2 S(4) emission.

5. Conclusions

We present new spectrally-resolved observations towards HH 54 of H2 0−0 S(4), 0−0 S(9), and 1−0 S(1). These are complemented by new Herschel/HIFI H2O (212−101) observations. Our data show for the first time the separation in velocity between the gas component traced by the low-excitation H2 0−0 S(4) line and that associated with the H2 lines at higher excitation. The observed H2 stratification in velocity suggests that our observations resolve two distinct gas components associated with the HH 54 shock region at different velocity and excitation. We spectrally identify the H2 0−0 S(4) line as the H2 counterpart of H2O emission. This allows us to constrain the temperature of the H2O emitting gas (≤1000 K). H2O abundance is estimated to be lower than what is expected from shock model predictions by at least one order of magnitude. High spectral resolution observations of different targets are needed to confirm this result.

2

The data are part of the OT2 program “Herschel observations of the shocked gas in HH 54” (observation ID: 1342251604).

Acknowledgments

We thank VLT astronomers and operators for performing excellent service mode observations at CRIRES and providing excellent support with VISIR. We particularly thank the ESO Director’s Office for the DDT observations with CRIRES. This work was partly supported by ASI–INAF project 01/005/11/0, PRIN INAF 2012 – JEDI, and Italian Ministero dell’Istruzione, Università e Ricerca through the grant Progetti Premiali 2012 – iALMA.

References

All Tables

Table 1

H2 transitions observed.

In the text

All Figures

thumbnail Fig. 1

Upper: H2 1−0 S(1) image of HH 54 from NTT/SofI observations (Giannini et al. 2006). The positioning of the slits adopted for VLT/VISIR and CRIRES observations is shown in blue and red. The beam sizes of Herschel H2O (212−101) (magenta circle), CO (15−14) (green, Bjerkeli et al. 2014), and SEST CO (2−1) (black, Bjerkeli et al. 2009, 2011) are displayed. Offsets are relative to: αJ2000 = 12h55m503, δJ2000 = − 76°56′23′′ (Bjerkeli et al. 2011). Lower: Spitzer/IRS H2 0−0 S(4) image of HH 54 (Neufeld et al. 2006). Symbols are the same as in the upper panel.
Open with DEXTER
In the text
thumbnail Fig. 2

Upper: H2 0−0 S(4) towards HH 54C1/C2 is compared with 1−0 S(1) and 0−0 S(9). Spectra are normalized to their peak values. The vertical dashed line marks the systemic velocity (vLSR = + 2.4 km s-1, Bjerkeli et al. 2011). The two vertical dotted lines indicate the velocity of: the 0−0 S(4) peak at − 7 km s-1; and the 1−0 S(1) and 0−0 S(9) peaks at −15 km s-1. Lower: H2O (212 − 101), CO (15−14), and CO (2−1) towards HH 54B are compared with H2 0−0 S(4) at HH 54C1/C2. H2O, CO, and H2 are normalized to the peak of the bump feature in CO (2−1).
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.