A&A 453, L47-L50 (2006)
LETTER TO THE EDITOR
K. M. Pontoppidan1,2
1 - California Institute of Technology, Division for Geological and Planetary Sciences, MS 150-21, Pasadena, CA 91125, USA
2 - Leiden Observatory, PO Box 9513, 2300 RA, Leiden, The Netherlands
Received 9 May 2006 / Accepted 22 May 2006
Aims. Ices in dense star-forming cores contain the bulk of volatile molecules apart from H2 and thus represent a large fraction of dark cloud chemistry budget. Mm observations of gas provide indirect evidence for significant freeze-out of CO in the densest cores. To directly constrain the freeze-out profile of CO, the formation route of CO2 and the carrier of the 6.8 m band, the spatial distribution of the CO/CO2 ice system and the 6.8 m band carrier are measured in a nearby dense core.
Methods. VLT-ISAAC, ISOCAM-CVF and Spitzer-IRS archival mid-infrared (3-20 m) spectroscopy of young stellar objects is used to construct a map of the abundances of CO and CO2 ices in the Oph-F star-forming core, probing core radii from to AU or densities from to with a resolution of AU.
Results. The line-of-sight averaged abundances relative to water ice of both CO and CO2 ices increase monotonously with decreasing distance to the core center. The map traces the shape of the CO abundance profile between freeze-out ratios of 5-60%. and shows that the CO2 ice abundance increases by a factor of 2 as the CO freezes out. It is suggested that this indicates a formation route of CO2 on a CO ice surface to produce a CO2 component dilute in CO ice, in addition to a fraction of the CO2 formed at lower densities along with the water ice mantle. It is predicted that the CO2 bending mode band profile should reflect a high CO:CO2 number ratio in the densest parts of dark clouds. In contrast to CO and CO2, the abundance of the carrier of the 6.8 m band remains relatively constant throughout the core. A simple freeze-out model of the CO abundance profile is used to estimate the binding energy of CO on a CO ice surface to .
Key words: astrochemistry - molecular processes - ISM: molecules - infrared: ISM
Theory has long predicted that molecules freeze out onto dust grains in dense molecular clouds causing gas-phase abundances to drop by orders of magnitude. A series of recent measurements of the gas-phase abundances of volatile molecules such as CO and N2H+in dense cores have corroborated this conjecture (e.g. Tafalla et al. 2004; Bacmann et al. 2002; Caselli et al. 1999). CO ice has also been observed directly along a growing sample of isolated lines of sight toward both embedded young stellar objects as well as background stars, showing that the CO molecules taken from the gas-phase re-appears in the solid phase (Pontoppidan et al. 2003; Chiar et al. 1994). CO ice is an excellent tracer of freeze-out processes because it is the only abundant ice species known to form initially in the gas-phase before adsorbing to a grain surface.
Pontoppidan et al. (2004) introduced the technique of ice mapping at high ( AU) spatial resolution - comparable to that of gas-phase maps. The procedure was demonstrated by constructing a map of the distribution of water and methanol ices in the outer envelope of the class 0 protostar SMM 4 in the Serpens star-forming cloud. The ice map of SMM 4 was used to show that the water ice abundace remained constant over a relatively wide range of densities, while the methanol abundance increased sharply by at least a factor of ten within 10 000 AU of the center of SMM 4. In principle, ices are best mapped toward field stars located behind the cloud. However, such background stars are typically extremely faint in the mid-infrared wavelength region. Therefore, disk sources embedded in their parent molecular cloud core offer a convenient infrared continuum against which the ices in the core can be mapped, with appropriate caveats on the interpretation of the derived ice abundances.
Direct mapping of CO ice abundances is highly complementary to mapping of gas phase abundances, because ice maps are sensitive to depletion fractions as low as 5%, while gas-phase abundance maps trace high depletion factors ( ) of (Caselli et al. 1999). Furthermore, the depletion fraction as a function of gas density in a core is strongly dependent on the properties of the dust grain surfaces. Direct maps of CO ice in a dense core therefore enables an independent measure of surface binding energies.
This letter presents the first spatial map of CO and CO2 ices in a dense molecular core. It will be demonstrated
that by directly measuring the relation between ice abundances and gas densities, the following quantities can be constrained: 1)
the dependence of CO depletion on gas density with a large dynamical range, 2) the CO-CO binding energy, 3) the formation route
of CO2 and the 6.8 m band carrier.
The case study is the F core in the Ophiuchus molecular cloud complex (Motte et al. 1998). Embedded within this region are at least
8-10 infrared-bright young stars. The lines of sight toward 5 of these young stars are used to probe the ices
in front of each star in order to obtain a profile of ice abundances across the minor axis of the core.
|Figure 1: Overview of the H2O, CO, 6.8 m and CO2 ice bands ( ordered left to right) used to construct the ice map of the Oph-F core. The 3.08 m water ice bands shown on the left-most panel are ordered according to optical depth, for clarity, as indicated in the panel. In the remaining panels, the sources are ordered according to projected distance to the core center, with the source furthest from the center (IRS 46) at the top. No. 5-8 m spectrum is available for IRS 46. Note also that the 6.8 m band of IRS 42 is filled in by emission likely due to PAHs. All the spectra have been shifted vertically, for clarity.|
|Open with DEXTER|
The individual spectra are taken from van Dishoeck et al. (2003) and Pontoppidan et al. (2003) of the 3.08 m and 4.67 m stretching modes of solid H2O and CO, obtained with the Infrared Spectrometer And Array Camera (ISAAC) on the Very Large Telescope (VLT). The solid CO2 component is probed along the same lines of sight using the InfraRed Spectrograph (IRS) on the Spitzer Space Telescope (AOR IDs 0009346048 and 0009829888 ) (Lahuis et al. 2006; Pontoppidan et al. 2005) as well as 5-16.3 m spectra obtained with ISOCAM-CVF (TDTs 29601715 and 29601813) (Alexander et al. 2003). The Spitzer spectra were taken as part of the Cores to Disks Legacy program (Evans et al. 2003). Combining these facilities, high quality 3-16 m spectra are available for 5 sources within a radius of 15 000 AU from the center of the pre-stellar core, defined to be the 850 m peak of Oph-F MM2 ( 27 24.3, 40 35, J2000) (Motte et al. 1998). Only IRS 46 has no suitable spectrum between 5 and 10 m.
To determine the optical depths of the various ice bands continua were fitted using low-order polynomials. The CO and CO2 bands are sufficiently narrow to make the continuum determination relatively straight-forward and unbiased. For the 3.08 m water ice band a continuum is fitted to points between 3.8 to 4.0 m and a K-band photometric point from the 2MASS catalogue. Finally, a local continuum between 5.5 and 8.0 m is used to extract optical depth spectra of the 6.8 m band. In consideration of the uncertainties in the continuum determination, care was taken to use the same "ice-free'' regions and second-order polynomials for all sources. Further discussion of the determination of continua for extracting ice optical depth spectra can be found in e.g. Gerakines et al. (1995); Dartois et al. (2002) and Keane et al. (2001). The ice spectra are shown on an optical depth scale in Fig. 1, while the locations of the sources relative to an 850 m JCMT-SCUBA map obtained by the COMPLETE collaboration (Ridge et al. 2006) are shown in Fig. 2.
The CO ice column densities have been determined in Pontoppidan et al. (2003). For the water 3.08 m band and the CO2 15.2 m band, band strengths of and are used, respectively (Gerakines et al. 1995). For the 6.8 m band, a band strength of is assumed, appropriate for NH4+ (Schutte & Khanna 2003). The ice column densities are summarized in Table 1.
|Figure 2: The locations (triangles) of the sources used to construct the ice map plotted on the SCUBA 850 m map. The star symbol indicates the center of the core.|
|Open with DEXTER|
Table 1: Line-of-sight averaged ice abundances in the Oph-F core.
It is seen that the abundances of pure CO,
water-rich CO and CO2 all increase toward the center of the core, while the 6.8 m band carrier decreases slightly in abundance.
The sharp rise in CO ice abundance is expected for CO freezing
out from the gas-phase at high densities in the central parts of the core.
Assuming a constant water ice abundance of
relative to H2 in accordance
with a model of the CRBR 2422.8-3423 line of sight of Pontoppidan et al. (2005), the average abundance of pure CO ice is seen to rise from
and the total CO and CO2 abundances from
This corresponds to a total CO depletion ranging from 12% to 60%, assuming an initial CO gas-phase
|Figure 3: Radial map of CO and CO2 ices of the Oph-F core. Ice abundances relative to H2O ice toward young stars are plotted as functions of projected distance to the center of the core as described in the text. The CO ice has been split into a "pure'' and "water-rich'' component following Pontoppidan et al. (2003).|
|Open with DEXTER|
The increase by roughly a factor two in CO2 ice abundance is particularly interesting. Since CO2 molecules are formed on the grain surfaces, the increase in CO2 ice abundance along with the CO freeze out is a direct indication of a formation route associated with CO. A significant amount of CO2 is present in the outer parts of the Oph-F core where the densities are not yet high enough for the CO to freeze out. This can be interpreted as evidence for two distinct eras for CO2 ice formation: The first takes place at roughly the same time as the bulk of the water ice is formed, which is known to happen as soon as the extinction, AV, into the cloud reaches a specific, relatively low, threshold of roughly 3-5 magnitudes (Whittet et al. 1988). This domain is not probed by the Oph-F ice map. The second takes place during the catastrophic freeze-out of CO that occurs at densities of a few (Jørgensen et al. 2005) in which an almost pure CO ice mantle forms. This second phase of CO2 molecules should be easily detectable in the shape of the 15.2 m CO2 bending mode absorption band, which is highly sensitive to the molecular environment of the CO2 molecules. Specifically, the new CO2 molecules should be found to be dilute in the CO ice with an average fraction of 1:5, as determined by the relative CO and CO2 abundance increases in the Oph-F core. Since the CO freeze-out rate is a very strongly increasing function of density, it is expected that the CO2:CO ratio will decrease in denser parts of the core. Thus, one would expect to find components with a range of CO2:CO ratios in the CO2 bending mode in the centers of dense clouds. The spectral resolution of the ISOCAM-CVF spectra ( ) is not sufficiently high to search for such band shape changes in the current data.
Since the young stars that are used to estimate ice abundances in the core
material are embedded in the core, significant caveats apply. Heating of the core material may desorb CO ice within a radius of a
few hundred AU of each source, depending on the luminosity. However,
since the projected size of the core is 30 000 AU, desorption is likely to play only a minor role. A possible exception
to this is IRS 46, which contains a significant component of warm molecular gas along the line of sight (Lahuis et al. 2006).
Also, the sources may be surrounded by remnant envelopes with different ice abundances than the surrounding core.
The ice absorption toward CRBR 2422.8-3423 was shown by Pontoppidan et al. (2005) to be dominated by cold core material, although a fraction
of the water, CO2 and 6.8 m ices are likely located in the disk. However, it
was found that the CO ice observed in this line of sight could not originate in the disk.
|Figure 4: Observed and modeled CO ice abundance for Oph-F. The curve shows the simple freeze-out model for the best-fitting CO binding energy of dH/k=814 K. The "solid CO'' abundance is the sum of the pure CO, CO in water and the CO2 ice excess abundances. This assumes that one CO molecule is used to produce each CO2 molecule. The error bars indicate a 20% uncertainty.|
|Open with DEXTER|
The CO depletion profile in Fig. 3 is modeled using a simplified freeze-out model, assuming a static core and
that only thermal desorption occurs. For the purposes of this letter, the model presented is intended as a demonstration of
the method, rather than achieving very accurate results. The rate of ice mantle build-up is then given by:
The radial map constitutes the first direct observation of the freeze-out profile of CO on dust grains in a prestellar core previously inferred indirectly from observations of millimetre lines of molecules. Additionally the observed increase in CO2 ice abundance toward the center is the first quantitative observational evidence of the formation of CO2 from CO on the surfaces of dust grains. These are observations that would not have been possible with single lines of sight.
This work was supported by a Spinoza grant. The author gratefully acknowledges discussions with Helen Fraser and Ewine van Dishoeck. The 850 m map of Oph-F from the COMPLETE survey was supplied by Doug Johnstone.