Edge-On Disk Study (EODS) III: Molecular Stratification in the Flying Saucer Disk

TL;DR

This study uses high-resolution ALMA tomography on the edge-on Flying Saucer disk to map the vertical and radial distribution of a broad molecular inventory, combining tomographic reconstructions with DiskFit radiative-transfer modeling. It demonstrates a common molecular layer whose altitude and thickness can be quantified across radii, with CN tracing the layer’s upper boundary and deuterated species like DCN and N$_2$D$^+$ residing nearer the mid-plane; most transitions are thermalized at $T_{ex}$ around $17-20$ K, while CO sits at higher altitudes. The results reveal a roughly constant molecular-layer temperature and a layered chemical structure consistent with astrochemical predictions, plus possible CO snowline proxying by N$_2$D$^+$. This direct measurement of molecular stratification provides a crucial benchmark for chemical modeling of protoplanetary disks and validates the relevance of edge-on systems for deciphering vertical disk structure.

Abstract

Context: Investigating the vertical distribution of molecular content in protoplanetary disks remains difficult in most disks mildly inclined along the line of sight. In contrast, edge-on disks provide a direct (tomographic) view of the 2D molecular brightness. Aims: We study the radial and vertical molecular distribution as well as the gas temperature and density by observing the Keplerian edge-on disk surrounding the Flying Saucer, a Class II object located in Ophiuchus. Methods: We use new and archival ALMA data to perform a tomography of $^{12}$CO, $^{13}$CO, C$^{18}$O, CN, HCN, CS, H$_2$CO, c-C$_3$H$_2$, N$_2$D$^+$, DCN and $^{13}$CS. We analyze molecular tomographies and model data using the radiative transfer code DiskFit. Results: We directly measure the altitude above the mid-plane for each observed species. For the first time, we unambiguously demonstrate the presence of a common molecular layer and measure its thickness: most molecules are located at the same altitude versus radius. Beyond CO, as predicted by chemical models, the CN emission traces the upper boundary of the molecular layer, whereas the deuterated species (DCN and N2D+) resides below one scale-height. Our best fits from DiskFit show that most observed transitions in the molecular layer are thermalized because their excitation temperature is the same, around 17-20 K. Conclusions: These long-integration observations clearly reveal a molecular layer predominantly located around 1-2 scale height, at a temperature above the CO freeze-out temperature. The deuterated molecules are closer to the mid-plane and N2D+ may be a good proxy for the CO snowline. Some molecules, such as CN and H2CO, are likely influenced by the disk environment, at least beyond the mm dust disk radius. The direct observation of the molecular stratification opens the door to detailed chemical modeling in this disk which appears representative of T Tauri disks.

Figures (9)

• Figure 1: Tomographic view of the observed spectral lines, showing the mean brightness as a function of radius and height (both in au). Original beam sizes are indicated in upper right corner of each panel. CN hyperfine transitions are labeled by their frequency in MHz. Contour levels are 0.5 to 2 by 0.5, 3 to 10 by 1, and 12 to 30 by 2 K (except for the weak para-H$_2$CO line, contour at 0.25 K only). The panel labelled "CN Singles" shows a weighted average of 3 unblended lines are at 338516, 340008 and 340019 MHz, shown with a contour step of 0.25 K. The continuum at 230 GHz is shown in the bottom left panel, with contour levels 0.1 0.2 0.4 0.8 and 1.2 K.
• Figure 2: $^{13}$CO tomography in false color with superimposition of the tomographies, in contours, of the HCN 4-3, CS 7-6, CN at 340.348 GHz, p-H$_2$CO at 218.222 GHz, DCN 3-2 and N$_2$D$^+$ 3-2. Contours are in steps of 1.5 K for HCN, CS, CN and p-H$_2$CO, 0.75 K for DCN, and 0.375 K for N$_2$D$^+$. The grey ellipses indicate the impact of line width on the effective beam at 80 and 240 au
• Figure 3: Top: Altitude $A(r)$ of the molecular layer as a function of radius. The black curve is the HCN(4-3) altitude for comparison. The cyan region indicates the the approximate size of the dust disk. The dotted line corresponds to z/r = 0.2. Bottom: Deconvolved thickness of the molecular layer. The deconvolution is done assuming Gaussian shapes, using the Clean beam minor axis since synthesized beams are elongated almost parallel to the disk plane. The black curve is the HCN(4-3) thickness for comparison. The cyan bar marks the edge of the dust disk.
• Figure 4: Profiles resulting from DiskFit at 100 au. The green curve is the temperature profile, and the cyan curves indicate the H$_2$ density structure (thin: under isothermal assumption at the mid-plane temperature (10 K), thick: using the full temperature profile) that were derived from the CO analysis by Guilloteau+2025. The other colored curves are the molecular density profiles for CS (red), HCN (black), CN 2-1 (blue) and 3-2 (magenta), using parameters from Table \ref{['tab:models']}. The corresponding horizontal bars indicate the measured temperatures and the density weighted average height of the emission.
• Figure 5: Tomographies of the residuals of the DiskFit models. For each line, the best fit model visibilities are subtracted from the observed ones, and imaged like for observations. The tomographic method is then applied on these residual data cubes. Contour spacings are 0.5 K.