ICES in the edge-on disk CRBR 2422.8-3423: Spitzer spectroscopy and Monte Carlo radiative transfer modeling

We present 5.2-37.2 mu m spectroscopy of the edge-on circumstellar disk CRBR 2422.8-3423 obtained using the Infrared Spectrograph (IRS) of the Spitzer Space Telescope. The IRS spectrum is combined with ground-based 3-5 mu m spectroscopy to obtain a complete inventory of solid-state material present along the line of sight toward the source. Archival JHKs imaging, as well as 350 mu m Caltech Submillimeter Observatory mapping, 850 mu m SCUBA mapping, and 3 mm Owens Valley Radio Observatory interferometry, is used to obtain a set of spectrophotometric data covering 1.2-3000 mu m. The ices observed toward CRBR 2422.8-3423 are compared with archival ISOCAM- CVF 5-16 mu m ice spectra of other nearby sources within 2'. We model the object with a two-dimensional axisymmetric (effectively three-dimensional) Monte Carlo radiative transfer code using all the available observations to constrain the source geometry and dust composition. In particular, the location of the observed ices in the disk and envelope material is included in the model. It is found that the model disk, assuming a standard flaring structure, is too warm to contain the very large observed column density of pure CO ice but is possibly responsible for up to 50% of the water, CO2, and minor ice species. In particular, the 6.85 mu m band, tentatively ascribed to NH4+, exhibits a prominent red wing, indicating a significant contribution from warm ice in the disk. The shape of the CO2 bending mode suggests an interaction with up to 20% of the CO ice. It is argued that the pure CO ice is located in the dense core Oph-F in front of the source seen in the submillimeter imaging, with the CO gas in the core highly depleted. Up to 50% of the CO ice embedded in water or CO2 ice (no more than 20% of the total amount of CO) may still be located in the disk, assuming constant abundances of these types of CO ice throughout the system. Discrepancies among the strength of different water ice bands are discussed. Specifically, the observed water ice libration band located at 11-13 mu m is significantly weaker than that of the model. The model is used to predict which circumstances are most favorable for direct observations of ices in edge-on circumstellar disks. Ice bands will in general be deepest for inclinations similar to the disk opening angle, i.e., similar to 70 degrees, except for very tenuous disks. Because of the high optical depths of typical disk midplanes, ice absorption bands will often probe warmer ice located in the upper layers of nearly edge-on disks. The ratios between different ice bands are found to vary by up to an order of magnitude depending on disk inclination because of radiative transfer effects caused by the two-dimensional structure of the disk. Ratios between ice bands of the same species can therefore be used to constrain the location of the ices in a circumstellar disk.