EDP Sciences
Herschel/HIFI: first science highlights
Free Access
Issue
A&A
Volume 521, October 2010
Herschel/HIFI: first science highlights
Article Number L6
Number of page(s) 5
Section Letters
DOI https://doi.org/10.1051/0004-6361/201015092
Published online 01 October 2010
A&A 521, L6 (2010)

Herschel/HIFI: first science highlights

LETTER TO THE EDITOR

A HIFI preview of warm molecular gas around χ Cygni: first detection of H2O emission toward an S-type AGB star[*]

K. Justtanont1 - L. Decin2,3 - F. L. Schöier1 - M. Maercker4,22 - H. Olofsson1,5 - V. Bujarrabal6 - A. P. Marston7 - D. Teyssier7 - J. Alcolea8 - J. Cernicharo9 - C. Dominik3,10 - A. de Koter3,11 - G. Melnick12 - K. Menten13 - D. Neufeld14 - P. Planesas6,15 - M. Schmidt16 - R. Szczerba16 - R. Waters3,2 - Th. de Graauw17 - N. Whyborn17 - T. Finn18 - F. Helmich19 - O. Siebertz20 - F. Schmülling20 - V. Ossenkopf20 - R. Lai21

1 - Onsala Space Observatory, Chalmers University of Technology, Dept. Radio & Spece Science, 439 92 Onsala, Sweden
2 - Instituut voor Sterrenkunde, Katholieke Universiteit Leuven, Celestijnenlaan 200D, 3001 Leuven, Belgium
3 - Sterrenkundig Instituut Anton Pannekoek, University of Amsterdam, Science Park 904, 1098 Amsterdam, The Netherlands
4 - University of Bonn, Argelander-Institut für Astronomie, Auf dem Hügel 71, 53121 Bonn, Germany
5 - Department of Astronomy, AlbaNova University Center, Stockholm University, 10691 Stockholm, Sweden
6 - Observatorio Astronómico Nacional. Ap 112, 28803 Alcalá de Henares, Spain
7 - European Space Astronomy Centre, ESA, PO Box 78, 28691 Villanueva de la Cañada, Madrid, Spain
8 - Observatorio Astronómico Nacional (IGN), Alfonso XII N$^{\circ}$3, 28014 Madrid, Spain
9 - CAB, INTA-CSIC, Ctra de Torrejón a Ajalvir, km 4, 28850 Torrejón de Ardoz, Madrid, Spain
10 - Department of Astrophysics/IMAPP, Radboud University Nijmegen, Nijmegen, The Netherlands
11 - Astronomical Institute, Utrecht University, Princetonplein 5, 3584 CC Utrecht, The Netherlands
12 - Harvard-Smithsonian Center for Astrophysics, Cambridge, MA 02138, USA
13 - Max-Planck-Institut für Radioastronomie, Auf dem Hügel 69, 53121 Bonn, Germany
14 - Johns Hopkins University, Baltimore, MD 21218, USA
15 - Joint ALMA Observatory, El Golf 40, Las Condes, Santiago, Chile
16 - N. Copernicus Astronomical Center, Rabianska 8, 87-100 Torun, Poland
17 - Atacama Large Millimeter/Submillimeter Array, Joint ALMA Office, Santiago, Chile
18 - Experimental Physics Dept., National University of Ireland Maynooth, Co. Kildare, Ireland
19 - SRON Netherlands Institute for Space Research, Landleven 12, 9747 AD Groningen, The Netherlands
20 - KOSMA, I. Physik. Institut, Universität zu Köln, Zülpicher Str. 77, 50937 Köln, Germany
21 - Northrop Grumman Aerospace Systems, 1 Space Park, Redondo Beach, CA 90278, USA
22 - European Southern Observatory, Karl Schwarzschild Str. 2, Garching bei München, Germany

Received 31 May 2010 / Accepted 9 June 2010

Abstract
Aims. A set of new, sensitive, and spectrally resolved, sub-millimeter line observations are used to probe the warm circumstellar gas around the S-type AGB star $\chi $ Cyg. The observed lines involve high rotational quantum numbers, which, combined with previously obtained lower-frequency data, make it possible to study in detail the chemical and physical properties of, essentially, the entire circumstellar envelope of $\chi $ Cyg.
Methods. The data were obtained using the HIFI instrument aboard Herschel, whose high spectral resolution provides valuable information about the line profiles. Detailed, non-LTE, radiative transfer modelling, including dust radiative transfer coupled with a dynamical model, has been performed to derive the temperature, density, and velocity structure of the circumstellar envelope.
Results. We report the first detection of circumstellar H2O rotational emission lines in an S-star. Using the high-J CO lines to derive the parameters for the circumstellar envelope, we modelled both the ortho- and para-H2O lines. Our modelling results are consistent with the velocity structure expected for a dust-driven wind. The derived total H2O abundance (relative to H2) is $(1.1 \pm 0.2)\times10^{-5}$, much lower than that in O-rich stars. The derived ortho-to-para ratio of $2.1\pm 0.6$ is close to the high-temperature equilibrium limit, consistent with H2O being formed in the photosphere.

Key words: stars: AGB and post-AGB - circumstellar matter - stars: kinematics and dynamics - stars: individual: $\chi $ Cyg - stars: late-type -
stars: mass-loss

1 Introduction

Observations of the dust and gas components in the circumstellar envelopes (CSEs) around asymptotic giant branch (AGB) stars have been carried out at different wavelengths. Observations in the infrared trace the dust as well as the warm molecular layer close to the stellar photosphere (e.g., Justtanont et al. 1996; Aoki et al. 1999; Schöier et al. 2002). Submillimeter and radio observations of trace molecules have been used to study the cooler outer parts of the CSEs (e.g., Schöier & Olofsson 2000; Kemper et al. 2003; Knapp & Morris 1985). To bridge this gap, the ISO was used to observe a large number of AGB stars up to almost 200 ${\mu}$m, and in O-rich stars, a number of H2O emission lines were detected (Barlow et al. 1996; Neufeld et al. 1996). However, the circumstellar far-infrared lines were unresolved. Hence, crucial information about the line profiles remained unknown.

Water is an important molecule in CSEs as it is thought to be one of the main cooling agents in the wind. It is also expected to be a good probe of the inner regions of the CSE where the gas is accelerated. However, to fully explore the potential of H2O lines as a probe of the circumstellar gas, a full radiative transfer has to be performed. Owing to difficulties in calculating accurate collisional rates coupled with the very high optical depth of the H2O lines in the inner region of the CSE, slow progress has been made. Nevertheless, calculations of the heating and cooling in the CSEs of O-rich stars suggest that H2O dominates the cooling in most parts of the envelope until it is photodissociated by interstellar UV photons (Goldreich & Scoville 1976; Maercker et al. 2009; Justtanont et al. 1994; Decin et al. 2010a; Maercker et al. 2008), and that some lines should come mainly from the acceleration zone. Eventually, spectrally resolved circumstellar H2O lines were observed by two space missions dedicated to search for cosmic water line emission: SWAS and Odin. Both missions were able to detect the ground-state line of H2O at 557 GHz in a number of AGB stars (Melnick et al. 2001; Harwit & Bergin 2002; Maercker et al. 2009; Justtanont et al. 2005; Hasegawa et al. 2006). It was shown that not only the line intensity, but also the line profile is crucial for interpreting the data correctly.

In 2009, the ESA-Herschel Space Observatory (Pilbratt et al. 2010) was launched with the Heterodyne Instrument for the Far-Infrared (HIFI[*], de Graauw et al. 2010), which aims to study H2O line emission in different environments in our Galaxy and beyond. HIFI offers the opportunity to study the warm molecular layers in CSEs of AGB stars in great detail, e.g., the high spectral resolution and wide spectral coverage allow a detailed study of the gas dynamics.

As part of the guaranteed time programme HIFISTARS (P.I.: V. Bujarrabal), the S-star (C/O $\approx$ 1) $\chi $ Cyg was selected for study. Distance estimates range from 150 pc (Knapp et al. 2003) to 180 pc (van Leeuwen 2007). The star exhibits SiO masers (e.g., Olofsson et al. 1981; Schwartz et al. 1982; Alcolea & Bujarrabal 1992), but no H2O maser emission has been found (Shintani et al. 2008; Menten & Young 1995). Being nearby and bright, $\chi $ Cyg has been observed using interferometric techniques in both the optical (Lacour et al. 2009) and the infrared (Tevousjan et al. 2004). In this Letter, we briefly describe the observations in Sect. 2, we discuss the modelling of the observed molecular emission in Sect. 3, and present our results and conclusions in Sect. 4.

Table 1:   HIFI observations of $\chi $ Cyg.

2 Observations

The HIFI data were obtained using the dual-beam-switching mode (Roelfsema et al. 2010) with a throw of 3$^{\prime}$ and slow (0.5-1 Hz) chopping in March-April 2010. A total of 8 frequency settings with a total of 10 lines detected are being reported in this paper. The targeted lines were selected to cover a wide range of excitation temperature, exploring different regions of the CSE. As backend, the wide band spectrometer (WBS) covering a region of 4 GHz with a resolution of 1.1 MHz was used. More details about these observations can be found in Bujarrabal et al. (2010). The data were calibrated using the standard pipeline for Herschel, HIPE version 2.8. Only the H-polarization data are presented here because the V-polarization data are noisier especially for the high frequency lines. We subtracted the baseline using a first or second order polynomial, except for the H2O transition at 1716 GHz, where this line is affected by standing waves and, consequently, the baseline was fitted using a high order polynomial (see Sect. 3.2).

We detected high rotational transitions of CO as well as the first detection of rotational H2O in an S-star. All of the observed lines listed in Table 1 were detected. The frequency $\nu$ in GHz and the energy of the upper level $E_{\rm up}$ in K are given along with the integrated line intensity, $I=\int\!\!T_{\rm mb}\rm d\upsilon$ in K km s-1. The spectra were corrected for the main beam efficiency, i.e., $\eta_{\rm mb} = 0.72~\rm exp(-(\nu(\rm GHz)/6000)^2)$. The absolute calibration accuracy ranges from 10% for the lowest frequency line to 30% for the high frequency (>1 THz) lines.

3 Modelling HIFI lines

We started the analysis by fitting the spectral energy distribution of the CSE (assumed to be spherically symmetric) using Dusty (Ivezic & Elitzur 1997) to derive the dust mass-loss rate. Based on this, we solved the gas-dust drag equation, assuming that both are momentum coupled, to derive the velocity structure using the observed terminal gas velocity as a constraint, as shown in Fig. 1. This also provides a so-called dynamical mass-loss rate estimate, $4.9\times 10^{-7}$ $M_{\odot}$ yr-1 in the case of $\chi $ Cyg (see e.g., Ramstedt et al. 2008). The gas velocity law is fed into the CO radiative transfer model, where the line fluxes and shapes are computed and compared to the observations. At the same time, the heating by dust grains and the cooling by CO lines are calculated and the resulting temperature structure obtained (e.g., Ramstedt et al. 2008; Decin et al. 2006; Justtanont et al. 1994; Crosas & Menten 1997; van der Tak et al. 2007). This provides a circumstellar model, including a mass-loss-rate estimate based on the CO line modelling, which is used to model the H2O line emission and its contribution to the cooling. The H2O cooling is then used to recalculate the gas kinetic temperature in the CO model, and the process is iterated until good fits to the observed CO and H2O lines are obtained.

\begin{figure}
\par\includegraphics[width=5.8cm,angle=-90,clip]{15092fg1.ps}
\end{figure} Figure 1:

The derived gas temperature structure (solid line) and the gas expansion velocity (dashed line) of the CSE of $\chi $ Cyg.

Open with DEXTER

3.1 Modelling of the CO lines

The Monte Carlo code developed by Schöier & Olofsson (2001) was used to model the observed CO lines. The molecular data were taken from the Hitran database (Rothman et al. 2009) and the collisional rate coefficients from Yang et al. (2010) for the 41 lowest rotational levels in the v=0 vibrational state. We fitted the line shapes and fluxes for the low-J lines obtained from ground-based observations, as well as the interferometric data for the J = 1-0 and 2-1 lines to more tightly constrain the size of the CO envelope (Schöier et al, in prep.). The parameters are listed in Table 2. The gas kinetic temperature distribution is shown in Fig. 1.

\begin{figure}
\par\includegraphics[width=12cm,clip]{15092fg2.eps} %
\end{figure} Figure 2:

The model fits (smooth lines) to the HIFI CO J = 6-5, 10-9 and 16-15 lines (histogram). The scaling factor (used to scale the model result to fit the observed integrated intensity) is given for each line, showing the goodness of the fit.

Open with DEXTER

Table 2:   Parameters used in the modelling of the line emission.

We assume a distance of 150 pc (Knapp et al. 2003), and the resulting mass-loss rate is $7~\times ~10^{-7}$ $M_{\odot}$ yr-1 using an adopted CO abundance of $6~\times~ 10^{-4}$ (relative to H2, see Table 2), which provide the best fits to both the line profile and line intensities. The uncertainty in the mass-loss rate is of the order 50%. This mass-loss rate agrees well with the dynamical mass-loss rate (certainly to within the errors in the input parameters). A comparison of the model fits and the HIFI observations can be seen in Fig. 2. The models have been scaled to match the line intensities and the scaling factor (given in each panel of the figure) is a measure of the goodness of fit. The high rotational line at J = 16-15 is noticeably narrower than the lower-level lines observed with HIFI and ground-based instruments (e.g., Knapp et al. 1998), indicating that this line originates in a region where the gas is still being accelerated. The estimated outer CO radius is $4~\times~ 10^{16}$ cm.

\begin{figure}
\par\includegraphics[width=17cm,clip]{15092fg3.eps} %
\end{figure} Figure 3:

The model fits (smooth lines) to the HIFI ortho- and para-H2O lines (histogram). The scaling factor (used to scale the model result to fit the observed integrated intensity) is given for each line, showing the goodness of the fit. The pipeline-reduced data of the 303-212 line are plotted along with the fitted baseline at the bottom right. This line is affected by standing waves within HIFI.

Open with DEXTER

3.2 Modelling of the H$_{\sf 2}$O lines

To calculate the strength and shape of the circumstellar H2O lines, we apply the parameters derived from our CO line modelling (mass-loss rate, gas temperature and density structure, and gas velocity law) to the radiative transfer model for H2O based on an accelerated lambda iteration (ALI) code (Maercker et al. 2009; Justtanont et al. 2005; Maercker et al. 2008). We used the molecular data from Rothman et al. (2009) and the collisional cross-sections from Faure et al. (2007) for the lowest 45 levels of ortho- and para-H2O. The radiative excitation due to the absorption in the $\nu_{2}$ bending and $\nu_{3}$ stretching modes is included. The latter has been found to have a non-negligible effect in the low mass-loss-rate case (Maercker et al. 2009). The outer radius of the H2O shell was derived using the model results of Netzer & Knapp (1987), i.e., $3.6~\times~ 10^{15}$ cm. Since the ortho- and para-species are expected to be independent, we model the two species separately, using the same circumstellar input values and estimate the two abundances independently. Using the temperature and velocity structure from the CO modelling (Fig. 1), we calculate the best-fit model to the H2O lines. All the lines are found to be sub-thermally excited. As mentioned above, both CO and H2O line cooling is taken into account.

We present the fits to four lines of ortho-H2O and three lines of para-H2O observed with HIFI in Fig. 3. These lines span upper energies from 60 to 300 K so the lines probe the cool part of the CSE as well as the warmer inner part. From Figs. 1 and 4, it can be seen that H2O lines originate well within the acceleration zone. All lines are of reasonably high signal-to-noise ratio, including the 303 - 212 line, the highest frequency line, where the spectrum is affected by standing waves inside HIFI. For this line, we used a higher order Chebyshev polynomial for the baseline subtraction (bottom right panel of Fig. 3).

Our results are consistent with the velocity structure of a dust-driven wind as can be seen in good fits to both the CO and H2O line profiles (Figs. 2 and 3, respectively). Unlike the case for IK Tau (Decin et al. 2010b,a), no modification to the dynamical calculation is required.

4 Results and conclusions

The observed HIFI lines are reliable probes of the inner CSE as the high-energy lines probe the wind in the acceleration zone. Both the velocity and density structures are tightly constrained using the HIFI lines. The abundance of CO used is $6\times 10^{-4}$, intermediate to those usually adopted for O- and C-rich CSEs. The derived ortho- and para-H2O abundances are significantly lower, $(7.5~\pm~1.4)\times 10^{-6}$ and $(3.6~\pm~0.5)\times 10^{-6}$, respectively (Table 2). These values are well below the limits for O-rich AGB stars of >10-4(Maercker et al. 2009; Justtanont et al. 2005; Maercker et al. 2008) consistent with $\chi $ Cyg being an S-star of C/O very close to unity. From our modelling, assuming that all carbon is locked up in CO (i.e., C/H = $3\times 10^{-4}$) and the oxygen is locked up in both CO and H2O (i.e., O/H = $3\times 10^{-4} + 5.5\times 10^{-6}$), our derived C/O is $\leq$0.98, given that a small fractional abundance of the oxygen is in dust grains. This value is slightly higher than that of 0.95, assumed by Duari & Hatchell (2000). A non-thermal equilibrium chemistry model for S-stars (C/O = 0.98) predicts an H2O abundance of 10-4 at the stellar photosphere, falling off to a few 10-6 at 5 R* (Cherchneff 2006), an order of magnitude lower than our value.

In the thermal equilibrium (TE) limit at high temperature, the expected ortho-to-para ratio is 3. Our derived ortho-to-para ratio is $2.1~\pm~0.6$, close to the high-temperature TE value. The reported ortho-to-para ratio in CSEs of O-rich stars vary from 1 in W Hya with a large uncertainty (Barlow et al. 1996) to 3 in IK Tau (Decin et al. 2010b). Our result is consistent with H2O molecules being formed under thermal equilibrium conditions in the warm and dense stellar photosphere.

\begin{figure}
\par\includegraphics[width=5.7cm,angle=-90]{15092fg4.ps} %
\end{figure} Figure 4:

The cooling rates due to different process : rotational cooling by H2O (solid), CO (dashed) lines, vibrational cooling by H2 (dotted) line, and adiabatic cooling (dot-dash).

Open with DEXTER

Given the low total H2O abundance of $(1.1~\pm~0.2)\times 10^{-5}$, it is clear from our analysis that the dominating cooling agent in the CSE of $\chi $ Cyg is CO (Fig. 4). Vibrationally excited H2 and rotationally excited H2O contribute only in the innermost (r<1015 cm) part of the CSE while CO line cooling extends further out until CO is photodissociated by the external UV field. Adiabatic cooling dominates only in the outermost part of the CSE. The derived H2O abundance, although much lower than in O-rich stars, is higher than that observed in C-stars, IRC+10216 (Melnick et al. 2001) and V Cyg (Neufeld et al. 2010), consistent with AGB stars of S-type being chemically intermediate between O-rich and C-rich AGB stars.

Acknowledgements
HCSS/HSpot/HIPE is a joint development (are joint developments) by the Herschel Science Ground Segment Consortium, consisting of ESA, the NASA Herschel Science Center, and the HIFI, PACS and SPIRE consortia. K.J., F.S., M.M., and H.O. acknowledge funding from the Swedish National Space Board. This work has been partially supported by the Spanish MICINN, within the program CONSOLIDER INGENIO 2010, under grant ``Molecular Astrophysics: The Herschel and Alma Era - ASTROMOL'' (Ref.: CSD2009-00038). R.Sz. and M.Sch. acknowledge support from grant N 203 393334 from Polish MNiSW. J.C. thanks funding from MICINN, grant AYA2009-07304. This research was performed, in part, through a JPL contract funded by the National Aeronautics and Space Administration.

References

Footnotes

... star[*]
Herschel is an ESA space observatory with science instruments provided by European-led Principal Investigator consortia and with important participation from NASA.
... (HIFI[*]
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 Astronómico Nacional (IGN), Centro de Astrobiología (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.

All Tables

Table 1:   HIFI observations of $\chi $ Cyg.

Table 2:   Parameters used in the modelling of the line emission.

All Figures

  \begin{figure}
\par\includegraphics[width=5.8cm,angle=-90,clip]{15092fg1.ps}
\end{figure} Figure 1:

The derived gas temperature structure (solid line) and the gas expansion velocity (dashed line) of the CSE of $\chi $ Cyg.

Open with DEXTER
In the text

  \begin{figure}
\par\includegraphics[width=12cm,clip]{15092fg2.eps} %
\end{figure} Figure 2:

The model fits (smooth lines) to the HIFI CO J = 6-5, 10-9 and 16-15 lines (histogram). The scaling factor (used to scale the model result to fit the observed integrated intensity) is given for each line, showing the goodness of the fit.

Open with DEXTER
In the text

  \begin{figure}
\par\includegraphics[width=17cm,clip]{15092fg3.eps} %
\end{figure} Figure 3:

The model fits (smooth lines) to the HIFI ortho- and para-H2O lines (histogram). The scaling factor (used to scale the model result to fit the observed integrated intensity) is given for each line, showing the goodness of the fit. The pipeline-reduced data of the 303-212 line are plotted along with the fitted baseline at the bottom right. This line is affected by standing waves within HIFI.

Open with DEXTER
In the text

  \begin{figure}
\par\includegraphics[width=5.7cm,angle=-90]{15092fg4.ps} %
\end{figure} Figure 4:

The cooling rates due to different process : rotational cooling by H2O (solid), CO (dashed) lines, vibrational cooling by H2 (dotted) line, and adiabatic cooling (dot-dash).

Open with DEXTER
In the text


Copyright ESO 2010

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.